The original of tliis book is in 
 tlie Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924029037468 
 
le Measurement of Attention 
 
 BY 
 L. R. GEISSLER 
 
The Measurement of Attention 
 
 BY 
 
 L. R. GEISSLER 
 
 THBSIS PRESENTED TO THE UNIVERSITY FAC- 
 ULTY OS" CORNELI, UNIVERSITY IN PARTIAL 
 FULFILMENT OF THE REQUIREMENT FOR 
 THE DEGREE OF DOCTOR OF PHILOSOPHY 
 
 Reprinted from the American Journal of Psychology 
 October, 1909, Vol. XX, pp. 473-529. 
 
THE MEASUREMENT OF ATTENTION^ 
 
 By I/. R. GbiSSLER 
 
 CONTENTS 
 
 I. Previous Views and Methods 473 
 
 i. Systematic 474 
 
 ii. Experimental 485 
 
 a. Peripheral Vision 493 
 
 b. Muscular Strength 494 
 
 c. Liminal and Differential Sensitivity 495 
 
 d. Reaction Time 496 
 
 e. Degree of Precision 497 
 
 f. Graded Distraction 501 
 II. Measurement of Attention in Terms of Clearness Values 502 
 
 i. Experiments on Motor Inhibition 503 
 ii. Experiments on Degree of Clearness: Continuous 
 
 Addition 509 
 iii. Experiments on Degree of Clearness: Marking 
 
 Circles 515 
 
 III. Summary of Results 529 
 
 PART I 
 
 A Criticai, Study of Previous Vibws and Methods 
 
 As soon as the experimental method was carried into psy- 
 chology and applied to mental phenomena, the fundamental 
 importance of attention, which until then had not been gen- 
 erally recognized,^ became convincingly manifest. There is, 
 indeed, no field of psychological experimentation in which 
 attention does not play a determining part, or, in Ribot's 
 words, "attention is, in fact, the fundamental psychical condi- 
 
 1 Prom the Psychological Laboratory of Cornell University. 
 
 2W. James: Princ. of Psychol. I, 1890, 402; Ebbinghaus: Grundz. 
 d. Psychol., 2d ed., 1905, 6io f.; Titchener: I<ect. on Feeling and At- 
 tention, 1908, 172-174. 
 
474 GEISSLER 
 
 tion of almost all psychometrical researches." ' Results are 
 widely different, according as they are or are not obtained 
 under maximal attention. Hence the precise distinction be- 
 tween various degrees of attention becomes a matter of prac- 
 tical necessity. As such it has made itself felt in two ways. 
 On the one hand, it is often desirable to know how certain 
 results are influenced by the degree of attention, and the ex- 
 perimenter has consequently to use some means of inducing 
 variations of attention, though he need not make any particu- 
 lar effort to measure them accurately. For such purposes two 
 general methods have been mainly employed, the one consisting 
 in distracting the observer from a certain task to which his 
 whole attention is required, and the other demanding that he 
 perform simultaneously two different mental or motor opera- 
 tions. On the other hand, certain investigations have been 
 undertaken with the express purpose of discovering reliable 
 methods and standards for the measurement of all possible 
 degrees from maximal to minimal attention. 
 
 Our study of previous views and methods regarding the 
 measurement of attention will naturally divide into three 
 parts. First we shall take up the theoretical discussions con- 
 cerning the nature of degrees of attention. In the second 
 place, we shall review the general experimental methods 
 usually employed for inducing different degrees of attention 
 without attempt at their accurate measurement. Finally, we 
 shall have to study the special methods and standards pro- 
 posed or actually used for measuring attention. This, in gen- 
 eral, is the programme which we have laid out for ourselves 
 in Part I of the present study. In the second Part we shall 
 submit experimental evidence for the possibility of an adequate 
 measurement of attention in terms of clearness values. 
 
 Among the German psychologists, probably the first sys- 
 tematic reference to attention was made by I/eibniz" in his 
 distinction between perception and apperception. The differ- 
 ence between these two states of consciousness, as Wundt ' 
 has pointed out, was simply a matter of degrees of clearness. 
 Wundt in accepting this important distinction seems almost 
 surprised to find ' 'how correctly I<eibniz had already observed 
 that there were no sudden or abrupt changes but rather gradual 
 transitions between the different clearness degrees of the ideas. "* 
 Such a view implies, on the one hand, that there can be no 
 absolute limit separating the low degrees of clearness included 
 
 iRibot: The Psychology of Attention, Transl., 1896, 67. 
 'The Philosophical Works of Leibnitz. Tr. by G. M. Duncan. The 
 Mouadology, paragraphs 19-25, esp. 24. 
 'Wundt: Vorlesungen, 2d ed., 1892, 262. 
 ^Op. cit., 262. 
 
THB MEASUREMENT OF ATTENTION 475 
 
 in perception from the higher degrees of apperception, and, on 
 the other hand, that there must exist an indefinitely large 
 number of such degrees. A somewhat later and more ex- 
 plicit emphasis upon clearness as an essential factor in atten- 
 tion is found in Wolff's writings. He gives, for example, the 
 following definition: "facultas efficiendi, ut in perceptione 
 composita partialis una majorem claritatem ceteris habeat, 
 dicitur Attentio." As an illustration he mentions the case of 
 a man listening to a conversation so attentively that he en- 
 tirely forgets his visual surroundings and does not notice the 
 temperature of the room until his attention is called to either 
 the one or the other, when he also loses the thread of the con- 
 versation. Many other passages of a similar character could 
 be quoted from Wolff. ^ However, the significant fact for us is 
 that, as Wundt says, "since Leibniz the distinction between 
 clear and obscure ideas has remained an almost uncontroverted 
 possession of psychology.'" Certainly it was quite familiar 
 to the German psychologists of the i8th century, as a few 
 references from Dessoir will suffice to show.' Sulzer (1720- 
 1779), for example, speaks of attention as "das Licht der 
 Seele" and points out "dass durch vermehrte Aufmerksam- 
 keit ein blosses Bild der Phantasie das Leben eines wirklichen 
 Gegenstandes bekommeu und hingegen der wirkliche Ein- 
 druck durch. den Mangel der Aufmerksamkeit zu einer sehr 
 schwachen Vorstellung werden kann." Hentsch (1723-1764) 
 says: "diese Bemiihung, auf einen Teil einer vorkommenden 
 Sache mehr Acht zu haben, als auf die iibrigen, nennen wir 
 die Aufmerksamkeit"; while von Irwing (1728-1801) thinks 
 that "die Klarheit und Lebhaftigkeit aber allein ist nicht hin- 
 langlich, um eine Idee bis zur Apperception zu bringen." 
 
 The psychological writings of a later period, especially at 
 the beginning of the 19th century, were too seriously in- 
 fluenced by the philosophical interests and theories of their 
 authors to give unbiassed consideration to our topic. At the 
 same time, we have already arrived at the ' 'turning-point of 
 modern psychology,"^ for in Herbart's system the problem of 
 apperception and attention begins to assume fundamental im- 
 portance. His discussion of the limen of consciousness and of 
 the rise and fall of competing ideas can easily be interpreted as 
 referring to the increase and decrease of the clearness of ideas. 
 This is particularly true of his Psychologische Unterstichung 
 uber die Starke einer gegebenen Vorstellung als Function ihrer 
 
 iPsychologia empirica, 1732, par. 237; cf. par. 235-238, and 592; 
 Psychologia rationalis, 1734, par. 357, 358, 360, 367, 369, 372. 
 ^Wundt: Vorlesungen, 2d ed., 262. 
 
 8 Dessoir: Gesch. d. neueren deutschen Psychologie, 1894, 235-237. 
 * Titchener: Exper. Psych., Vol. I, Part II, 186. 
 
476 GBISSLER 
 
 Dauer betrachtet} While Herbart was the first to apply 
 mathematical formulas to mental phenomena, he nevertheless 
 thought it impossible to perform experiments upon human be- 
 ings, and he thus implicitly denied that psychology could 
 become an exact science. Lotze, on the other hand, with his 
 expert medical knowledge and his deep psychological insight, 
 helped indirectly to affiliate psychology to the other natural 
 sciences. He seems to have been the first to introduce into 
 modern psychological literature the analogy of attention and 
 inattention to the Blickpunkt and Blickfeld of vision. He 
 says: "the mind is not so constituted as to experience all its 
 (simultaneous) contents with equal clearness and attention. 
 It is rather to be compared to the retina of the eye with its 
 singularly sensitive fovea around which a large area of sym- 
 metrically decreasing sensitivity is centred. Just as in the 
 eye every peripheral point, in spite of its indistinctness, has 
 nevertheless its definite position with regard to the more dis- 
 tinct centre, so in the course of ideas in the mind the more 
 suppressed excitations centre around the clear focus of atten- 
 tion and, without disturbing the focal content, contribute to the 
 greater richness of consciousness and to its peculiar moods or 
 states of illumination."^ IvOtze also speaks of a "Steigerung der 
 Aufmerksamkeit" which manifests itself in peculiar changes 
 of the "Helligkeit gewisser Sinneseindriicke,"'a phrase which 
 evidently refers to the clearness of the sensation, since the 
 intensity of the stimulus is assumed to be constant. 
 
 With the beginnings of . experimental psychology in the 
 middle of the 19th century, a new emphasis was laid upon 
 attention. Many years before Wundt had started his inves- 
 tigations with the complication pendulum and the tachisto- 
 scope, t^echner had been led by his quantitative study of the 
 psychophysical relation to distinguish sharply the limen of 
 sensible intensity from the limen as influenced by the degree 
 of attention. As early as i860 he wrote, for instance: "We 
 can think or have ideas, weak as to their content, but strong as 
 to their intensity. . . . Hence the intensity of the idea and the 
 strength with which we think or perceive it must be in some 
 way distinguishable from each other".* In 1877 ^^ ^^s still 
 more explicit on the same point: "If we perceive a sensory 
 phenomenon or represent it to ourselves in the form of an idea, 
 the intensity of our conscious activity is then determined on 
 the one hand by the degree of attention with which we perceive 
 
 1 Herbart' s Sammtliche Werke, ed. by Karl Kehrbacli, 1888, Vol. 
 Ill, 121-145, esp. 123. Cf. I/ehrbuch z. Psych., Werke VI, 140 ff. 
 'Lotze: Medicinische Psychologie 1852, 505. 
 8 Op cit., 510. 
 * Fechner: El. d. Psychophysik, 2d ed., Vol. II, 453. 
 
THE MEASUREMENT OF ATTENTION 477 
 
 the sensory or the memory image, and on the other hand by 
 the vividness or intensity which pertains to the phenomenon 
 itself. ... In such cases we can quite well distinguish how 
 much is due to the degree of attention and how much intensity 
 belongs to the phenomenon as such".' A very similar passage 
 occurs in the Revision: "If I look at a more or less bright 
 surface or listen to a more or less loud sound, then I can very 
 well distinguish the degree of attention with which I perceive it 
 from the degree of brightness or of intensity of the sound which 
 I experience' ' . ^ He even goes further and adds : ' 'the strength 
 of the attention may decrease while the intensity of the 
 phenomenon, as, e. g., the brightness or sound sensation, may 
 increase, and conversely".' Now what else could such re- 
 peated emphasis upon the distinction between degrees of 
 psychophysical intensity and degrees of attention indicate, if 
 Fechner did not wish to differentiate as sharply as possible 
 between intensity and clearness? Although it is true that he 
 himself nowhere uses the term clearness in connection with 
 degrees of attention, yet it is equally certain that there is no 
 other sensational attribute than that of clearness which could 
 be thus fundamentally opposed to the attribute of intensity. 
 It is perhaps due to Fechner's avoidance of the term clearness 
 that he did not state more explicitly the inevitable conclusion 
 implied in these passages, that the degree or strength of the 
 attention given to a certain mental process is equivalent to the 
 degree of clearness which this process attains as compared with 
 the other mental processes of the moment. 
 
 In this respect Stumpf seems to go beyond Fechner, for in 
 the second volume of the Tonpsychologie there is a passage 
 which indicates that for Stumpf clearness is not only essential 
 to attention, but also that it increases as attention itself in- 
 creases. He says : ' 'the effect of attention in analysing is 
 capable of an almost unlimited increase. An analysed clang 
 and a particular tone in it may in all inner characteristics and 
 external relations, in all respects and in all directions become 
 more and more distinct".* Here the terms "Steigerung" of 
 attention and "immer deutlicher werden" are plainly synony- 
 mous. With regard to the psychophysical basis of attention 
 Stumpf accepts and defends^ Fechner's analogy of the wave 
 form against G. %. Miiller's criticism,' and quotes Exner' in 
 his support. 
 
 iln Sachen d. Psychophysik, 1877, 85 f. 
 
 2 Revision d. Hanptpnnkte d. Psychoph., 1882, 270. 
 
 8 Op. cit., 270. 
 
 5 Tonpsych. I, 1883, 70, footnote. 
 
 * Stumpf: Tonpsychologie II, 1890, 288. 
 
 « Miiller: Grundl. d. Psychophysik, 1878, 359. 
 
 ' Pflviger's Archiv XI, 1875, 429- 
 
478 GBISSLBR 
 
 In Wpps' discussion of attention, its relation to clearness is 
 made still more explicit. The function of attention is to 
 "increase that energy of the excitations with which they are 
 to gain independent consciousness",^ and Lipps sharply dis- 
 tinguishes this increased mental energy from the stimulus 
 intensity. In another passage he gives an admirable descrip- 
 tion or definition of clearness : "it is indeed possible for ideas 
 to break their way into consciousness with greater or less 
 strength, to remain there a longer or shorter time, and 
 accordingly to make a more or less energetic impression, to 
 enter into more or less close relations with other mental con- 
 tents, to be more or less appropriated by the total conscious 
 situation and thus to occupy in it a more or less dominating 
 position". '^ The same thought occurs twenty years later in a 
 more concise form in his Leitfaden, where he says, for example : 
 "the real psychical fact which lies at the basis of attention 
 may be defined as a more or less strong psychical effectiveness 
 of the processes upon which the attention is directed",' or: 
 ' ' the clearness degree of a sensation or idea would then be only 
 another name for its degree of effectiveness in the mental 
 life".* "Instead of psychical effectiveness", lyipps writes 
 elsewhere, "we may just as well say: psychical force. We 
 speak of force when we refer to some effect or work done. 
 And we measure the magnitude of the force by its effects. 
 Thus the amount of attention directed to an ideated object is 
 nothing but the psychical force of the idea of this object".^ 
 This means in other words that, in order to measure the atten- 
 tion given to a certain mental process, we have to consider its 
 ' ' psychical effectiveness " , or better, the degree of clearness 
 with which it appears in consciousness. 
 
 Kiilpe agrees with Fechner, Stumpf, and Lipps, when he 
 writes: "It is a matter of familiar experience that the vivid- 
 ness or degree of attention is itself capable of quantitative 
 gradation, and that the differences may be distinguished from 
 differences in the intensity of sensations."' However, he is 
 very sceptical as to the fruitfulness of the experimental meth- 
 ods employed in securing and measuring degrees of attention, 
 and closes his discussion by saying: "The discovery of a re- 
 liable measure of the attention would appear to be one of the 
 most important problems that await solution by the experi- 
 mental psychology of the future".' Ten years later he prac- 
 
 'I/ipps: Grundtatsachen d. Seelenlebens, 1883, 134. 
 "^Op. cit.,dfl i. 
 
 • Leitfaden : 1903, 34. 
 4Leitf: 38. 
 
 ^ Leitfaden : 34. 
 
 * Kiilpe: Outl. of Psych., 1895, 441. 'Kiilpe: Outl., 429. 
 
THE MEASURBMENT OF ATTENTION 479 
 
 tically repeats this statement in the following passage: "It is 
 clear that, if we are to induce a determinate degree of con- 
 sciousness, we must know, first, the mode of distribution of 
 the total energy of attention over the separate conscious con- 
 tents included within its range, and, secondly, the distractive 
 value of the process which we have selected for our purpose. 
 Now no general rule can be laid down under either heading ; 
 the results that have been obtained hold only in particular 
 cases ; so that, from this point of view also, the investigation 
 of attention is peculiarly difficult. So much we may say, and 
 no more : that the distribution of attention brings with it a 
 reduction, and concentration of attention an enhancement, of 
 the degree of consciousness for a determinate content. Psy- 
 chologists have not yet attained to any quantitative formulation 
 of the changes".^ 
 
 A similar complaint was made in the same year, as well as 
 two years later, by Ebbinghaus: "A desirable aim for the 
 study of the distraction phenomena would be the securing of 
 exact and more or less generally valid figures. . . . How- 
 ever, it is clear that the prospects for accomplishing this aim 
 under the described complication of things are not very 
 bright".'' 
 
 It is difficult to state correctly Wundt's attitude toward the 
 problem of measurement of attention, since he makes no ex- 
 plicit references to it. As was said above, Wundt accepts 
 I^eibniz' distinction between clear and obscure ideas, between 
 apperception and perception,' and he also seems to agree with 
 Leibniz as to the possibly endless number of intermediate de- 
 grees between these two extremes. Since they are always 
 accompanied by feelings of strain, the intensity of the feelings 
 may, according to Wundt's earlier writings, serve as an indi- 
 cation of the probable degree of apperception or clearness. 
 There seems here to be an ambiguity in Wundt's terminology, 
 in as far as he uses synonymously Spannungsgefuhl and Span- 
 nungsempfindung. But since strain, according to him, con- 
 stitutes the main element of the 7 hatigkeitsgefuhl, .since in his 
 later pluralistic theory of the afiections it is one of the six 
 dimensions of feeling, and finally, since he himself objects very 
 strongly to Miinsterberg's wide- reaching use of the muscular 
 and other bodily strain sensations, it seems probable that even 
 in his earlier writings he intended to classify strain among the 
 feelings. It is then not surprising that in other psychological 
 treatises, referring to Wundt's earlier writings, the terms sen- 
 sation and feeling of strain are likewise confused. Mentz, for 
 
 1 The Problem of Attention, Monist XIII, 45. 
 
 2 Ebbinghaus : Grundz. d. Psych., 2d ed., 1905, 623 f. 
 ^ Wundt: Vorlesnngen, 2d ed., 262. 
 
480 GBISSLEE 
 
 example, uses as a basis for a number of experiments the fol- 
 lowing passage in Wundt : "Those strain sensations which, 
 with the same attention, accompany the external volitional 
 act as well as the direction of the will to the particular sense 
 departments, form a complex of qualitatively related sensa- 
 tions which is characterized by a constancy attained by no 
 other conscious content. These sensations are always present, 
 — which cannot be said of any other sensory process ; and 
 through their various degrees of intensity they offer at the 
 same time a direct measure of the energy of any given voli- 
 tional activity"/ Mentz, accordingly, gives his observers the 
 following instruction : "The reactor is required voluntarily to 
 increase his attention in equal degrees, as far as this is possi- 
 ble, and to report afterwards what particular degree he has 
 given", and goes on to say: "Wundt has already remarked 
 that the strength of attention may be measured by the intensity 
 of the strain sensations"." Again, Wundt says, after speak- 
 ing of the relation of association to active and passive apper- 
 ception, that "in both cases the apperception, as a process 
 which differs from the mere course of ideas, comes to conscious- 
 ness partly by means of the accompanying feelings and partly 
 by means of the accompanying strain sensations whose inten- 
 sity increases with the degree of attention".' This leads Pills- 
 bury to make the statement that, according to Wundt, ' 'the 
 feeling of activity and the sensations of strain are most inten- 
 sive when the ideas are clearest. Increase in the clearness of 
 ideas and in intensity of strain sensations, etc. , always go on 
 side by side".^ Hence Mentz was justified in his interpreta- 
 tion of Wundt, and 6n this basis his observers distinguished, 
 according to his tables, in the one case four and in the other 
 case six degrees of voluntary attention. Unfortunately, Mentz 
 does not give any further description of these degrees and 
 adds only incidentally the remark "that just this increase 
 (probably of strain) with the higher degrees was rather diffi- 
 cult to accomplish".* Wundt himself, in the 5th edition of 
 the Physiologische Psychologie, has made a sharp distinction be- 
 tween the feelings and the sensations of strain. He now con- 
 nects the feelings of strain with the intensity of the sensory 
 processes attended to, because they inform us of our correct 
 bodily accommodation to the external stimulus and thus indi- 
 rectly help us to bring it to its full intensity. He says : ' 'Fur- 
 thermore we notice that the degree of the feelings of strain 
 
 'System d. Philos., 1889, 100. 
 
 2 Mentz: Phil. Stud., XI, 1895, 574. 
 
 'Grundz. d. physiol. Psych., II, 4th ed., 1893, 279. 
 
 *Pillsbury: Amer.Jour. Psych., VIII, 1896, 279. 
 
 5 Meutz : op. cii., 576. 
 
THE MEASUREMENT OF ATTENTION 48 1 
 
 keeps step with the intensity {Starke) of the impressions. 
 Upon the exactness of the occurring accommodation depends 
 what is called the keenness {Schdrfe) of attention".^ By this 
 keenness he means simply the accuracy of the sense organ in- 
 volved. After all, then, we are still left in the dark as to Wundt's 
 attitude toward the problem of measuring attention. It seems 
 hardly possible that he would deny the existence of such a 
 problem, although his refusal to make any direct reference to 
 it or to the large literature in question might perhaps be in- 
 terpreted in this way. However that may be, this much is 
 certain : Wundt at least admits a large number of degrees of 
 attention and makes them equivalent to degrees of clearness 
 ranging all the way from the highest level of apperception to 
 the lowest level of perception. On this point he agrees with- 
 all the other German authors quoted thus far ; in fact, his in- 
 sistence upon clearness as the essential characteristic of atten- 
 tion has become the central aspect of his whole system of 
 psychology. 
 
 In Miinsterberg's treatment of attention great stress is laid 
 upon kinaesthetic strain sensations. They are said to arise not 
 only in the muscles involved in the adjustment of the particu- 
 lar sense-organ, but also in the muscles of the head, the neck, 
 and the chest, especially in the muscles involved in respiration. 
 Miinsterberg holds that "the greater the demands that are 
 made upon the efi&ciency of the sense organ, the more intensive 
 will all these strains have to be, for the adaptation (^Einstel- 
 lung) takes place the more exactly, the stronger the tension in 
 the antagonistic muscles. . . . The intensity of the strain 
 sensations is therefore generally accepted as a measure of the 
 strength of attention; thus the latter may be increased con- 
 siderably while the impression itself remains constant' ' . " Kiilpe, 
 however, makes the criticism, which is based upon introspec- 
 tive evidence, that "there is no simple or necessary propor- 
 tionality between the intensity of strain sensations and the 
 degree of attention".' This objection cannot be satisfactorily 
 answered, and therefore Miinsterberg's criterion for degrees 
 of attention is psychologically impossible. 
 
 The earliest and perhaps the.only systematic French treatise 
 on attention is written by Ribot. He also admits the existence 
 of many degrees of attention, and even attempts their graphical 
 representation. He says: "graphically we might represent 
 the totality of its normal and morbid manifestations by a 
 straight line. ... At the centre, let us put ordinary 
 
 1 Grundz. d. physiol. Psychol., 5th ed.. Ill, 1903, 337. 
 
 2 Miinsterberg : article on Aufmerksamkeit in the Real-lexikou der 
 Medic. Propadeutik by Joh. Gad, Vol. I, 1893, col. 532-539- 
 
 ' Kiilpe : Outlines, 428. 
 
482 GEISSLER 
 
 Spontaneous attention. Following our imaginary line to the 
 right, in the direction of increasing attention, we find strong 
 spontaneous attention, then preoccupation, then the weak 
 fixed idea. . . . Reverting to our starting point, we 
 now turn to the left, in the direction of decreasing intensity. 
 Here we have voluntary attention, at first in the form of an 
 organized habit, then in its general ordinary form, then vacil- 
 lating. . . . Between each form and its adjacent ones, 
 there occur shades which I omit to notice".^ As to the nature 
 of these "shades", Ribot is thus not very explicit; but we 
 can easily infer from the passage quoted as well as from many 
 other statements that he must have had in mind differences 
 in clearness. This is, for instance, evident from the following 
 statement: "I am, accordingly, fully inclined to hold, with 
 Buccola, that the fixed idea is attention at its highest degree — 
 the extreme limit of its power of inhibition"." 
 
 Much less unanimity of opinion both as to the number and 
 the nature of degrees of attention exists among the English 
 and American psychologists. Ward, for instance, seems to 
 agree with Ribot and the German school, although he, too, is 
 not very explicit. He distinguishes between "the intensity or 
 concentration of attention and its diffusion or the extent of the 
 field of consciousness".* Still, he makes "loss of intensity" 
 correspond to loss of "distinctness" in the following passage: 
 "if, then, attention be distributed over too wide a field, there 
 is a corresponding loss of intensity, and so of distinctness". 
 He also seems to imply a scale of gradual changes ' 'between 
 the two zeros of complete indifference and complete absorp- 
 tion' ' . Practically at the same time Sully published his first 
 works on psychology. As he was one of the first English 
 systematic writers to give attention an independent place, it is 
 interesting to find him also in agreement with the Germans so 
 far as degree of attention is concerned. He says, for instance : 
 "the field of consciousness, however, is wider than that of at- 
 tention. Consciousness admits of many degrees of distinct- 
 ness. ... To attend is to intensify consciousness by 
 concentrating or narrowing it on some definite and restricted 
 area. It is to force the mind or consciousness in a particular 
 direction so as to make the objects as distinct as possible".* 
 However, since he thinks it best "to reserve the term attention 
 for the more palpable exertions of mental activity in definite 
 directions' ' ,' his terminology is different from that of previously 
 
 1 Ribot : The Psychology of Attention, 104-105. 
 
 " Op. cit., 86. Other evidences are quoted in Tilchener's Lectures, 
 
 'Ward: art. on Psychology in the Encycl. Brit., 1886, 70. [184. 
 
 * Sully: Outlines of Psychology, 2d ed., 1885, 73. 
 
 ^Op. cit., 75, footnote. 
 
THE MEASUEEMBNT OF ATTENTION 483 
 
 mentioned writers. Especially the terms activity and energy 
 play an important part in his later treatments of attention, as 
 may be seen from the following passage: "the amount of at- 
 tention exerted at any time depends on two chief circumstances, 
 (a) the quantity of active energy disposable at the time ; (6) 
 the strength of the stimulus or force which excites the atten- 
 tion or rouses it to action. If there is great active energy a 
 feeble stimulus will snfiBce to bring about attention. The 
 healthy, vigorous child in the early part of the day has a 
 superabundance of energy which shows itself in attention to 
 small and comparatively uninteresting matters. On the other 
 hand, a tired or weakly child requires a proportionately power- 
 ful stimulus".^ Here the reference to clearness is replaced by 
 a strong emphasis upon the hypothetical mental energy. 
 
 Practically the same thing holds of I,add's discussion of the 
 nature of attentive degrees. He begins by following Wundt, 
 for example when he writes: "that changes in the clearness of 
 perception take place in dependence on the changes in the de- 
 gree of attention is a matter of the most ordinary experi- 
 ence,"^ or: "by gradual increase in the intensity of attention 
 (changing the casual glance into a steady look) these objects 
 become apperceived more and more clearly.'" However, he 
 adds to this plain statement of a psychological fact the physio- 
 logical hypothesis that ' 'under the influence of attention the 
 cerebrum has become more susceptible for certain impressions, 
 less so for certain others. Stored energy of the nerve cells is 
 being rapidly called forth. Concentrated voluntary attention 
 implies a large amount of work being done in the cerebral 
 hemisphere."* Hence he concludes in a somewhat puzzling 
 way that "by the amount and speed expended in attention we 
 measure in a large degree the extent and intensity of con- 
 sciousness."' According to I^add "primary attention, essen- 
 tially considered, is the variously related degrees of psychic 
 energy expended upon the different aspects, elements, and 
 objects, in the one field of consciousness."' The admission 
 "of an indefinite number of degrees of (conscious) energy" he 
 thinks leads one to ask, for instance, "how much can I feel 
 without losing consciousness?" or "how vivid can my mem- 
 ory-image or picture of imagination become as measuring the 
 utmost capacity for vividness which my conscious mental life 
 can display in this way?"' These questions again refer to at- 
 
 ^ 0/>. cit., 79. 
 
 21/add: El. of Physiol. Psych., 1889, 541. 
 
 <^0p. cit., 541. 
 
 4 Op. cit., 543- 
 
 "Op. cit., 544- 
 
 « Psych. Descr. and Explan., 1894, 74-75. ' Op. cit., 42. 
 
484 GEISSLER 
 
 tention as essentially characterized by clearness, and thus we 
 find I/add's discussion of attention wavering between two 
 difierent standpoints. 
 
 According to Baldwin, the whole area of consciousness may 
 be subdivided into five levels or degrees. The lowest is be- 
 yond consciousness; it is the physiological region of the Un- 
 conscious. Then follows the Subconscious, next Passive or 
 Diffused Consciousness, then Active Consciousness or Atten- 
 tion, and finally the level of Apperception.^ In another pas- 
 sage he likens attention to "the line of mental vision," or to 
 "the visual field in which objects are scattered, those being 
 most clearly seen which are in the line of direct vision or cen- 
 tre of the field. Between these limits," he immediately 
 adds, "there are all degrees of distinctness."^ This last 
 sentence would seem to imply that each of the four or five 
 levels must itself include a large number of lesser differences, 
 although perhaps no hard and fast distinction could be drawn 
 between them. As to the nature of these degrees we are left 
 to infer that they are differences of clearness. Such an impli- 
 cation is for instance made in his discussion of the relation of 
 attention to the intellect, where he says: "in general, it may 
 be said that attention increases the vividness of representative 
 states,"' and lower down on the same page: "the capacity to 
 retain mental pictures depends upon the intensity of the origi- 
 nal presentation, and the clearness of its relations; and this 
 intensity and clearness are enhanced by the attention." How- 
 ever, in Baldwin's later writings the clearness factor is en- 
 tirely disregarded, a fact due probably to the dynamogenetic 
 point of view. In this respect he is not followed by J. R. 
 Angell, who otherwise seems to accept Baldwin's view as ex- 
 pressed in his earlier discussion of attention, even reprint- 
 ing the diagram of the five concentric circles which represent 
 the different levels of consciousness.* 
 
 The latest systematic treatise on attention, that of Pillsbury, 
 again emphasizes the clearness factor. He admits various de- 
 grees of attention, and identifies them with degrees of clear- 
 ness: "increase in the degree to which an impression is con- 
 scious and increase in attention to that impression are synony- 
 mous."* In his sixth chapter he discusses in some detail 
 "the Methods of Measuring Attention" and makes special 
 reference to the experimental work on this problem, which we, 
 too, shall have to consider later on. For the sake of com- 
 
 1 Baldwin: Senses and Intellect, 1889, 68. 
 
 ^Op.cit.,(>^. 
 
 'Baldwin: op. cit., 75. 
 
 * Angell: Psychology, 1904, 65-67. 
 
 * Pillsbury: Attention, 1908, 2. 
 
THB MBASURBMENT OF ATTENTION 485 
 
 pleteness we should mention in this place Titchener's latest 
 discussion of our topic, but since it also is intimately con- 
 nected with the experimental work upon attention we reserve 
 it likewise for later treatment. 
 
 To sum up our review of the theoretical discussions con- 
 cerning the nature and number of degrees of attention, we 
 find that most of the writers, as far as they are explicit on this 
 point, agree that there are an indefinitely large number of such 
 degrees, and that the degrees consist in differences of clear- 
 ness of the conscious processes attended to or distracted from. 
 
 We are now ready to pass to the experimental work upon 
 different degrees of attention. Here we have, first of all, to 
 distinguish between the general means of inducing variations 
 of attention on the one hand, and the special methods proposed 
 or used for the sake of standardising or accurately measuring 
 these variations on the other hand. Of course, all the inves- 
 tigations here in question had to employ some means of varying 
 the attention; but some of them did not go beyond this, simply 
 because their main problem lay elsewhere. Nevertheless, in 
 our present discussion of the various means of inducing differ- 
 ent degrees of attention, we shall not confine ourselves to this 
 class, but shall include also those of the other class, because 
 this procedure will enable us to reduce unavoidable repetition 
 of references to a minimum. I^ater on, when we come to dis- 
 cuss the special methods of standardising degrees of attention, 
 the investigations of the first kind will naturally have to be 
 omitted. 
 
 The main question for any one who attempts to vary atten- 
 tion is this : How can an observer be induced (or forced against 
 his will) to give less than maximal attention to a prescribed 
 task? The choice actually made among the numerous pos- 
 sibilities that offer themselves has depended partly upon the 
 nature of the given task, and partly upon the main purpose of 
 the particular investigation. Some of the authors have found 
 it necessary to use various means. Nevertheless, in general, 
 all procedures may be divided, as has been said in the intro- 
 duction, into two great classes, which we may briefly charac- 
 terise as the "single-task" method and the "double- task" 
 method, sometimes called the method of distraction and the 
 method of simultaneous activities. These two methods are 
 intimately connected with the basal fact of consciousness that 
 at any given moment only a limited number of mental processes 
 may be attended to, that is, may rise to focal clearness, while 
 other simultaneously present processes occupy the relatively 
 obscure background. In the single-task method, as the name 
 implies, either the processes constituting the prescribed task 
 
486 GEISSLER 
 
 are to be attended to, while other processes, constituting what 
 is usually called the distraction, are artificially introduced in 
 the background in order to reduce indirectly by their strength 
 and number the attention given to the focal processes of the 
 task ; or, more frequently (and especially in the later experi- 
 ments), the mental processes of the distraction are made focal 
 while the original task is relegated to the background. In the 
 double-task method, on the other hand, two sets of mental 
 processes are introduced which are intended to occupy simul- 
 taneously the focus of consciousness, while the background 
 processes are not taken into account. Kiilpe was the first to 
 suggest this distinction, when he said : ' ' experience shows that 
 there are but two ways, in the normal waking state, by which 
 this end may be accomplished. We may either distract the 
 attention, or we may divide it".^ In both cases, the variation 
 of attention is inferred from the change in quality or quantity 
 (or both) of the task performed.'' 
 
 The single-task method is, so far as our knowledge goes, 
 chronologically the older of the two. It seems to have been 
 first employed by Wundt' and Obersteiner*, as early as 1874. 
 Obersteiner, who attempted to measure attention by means of 
 the reaction time, used auditory, cutaneous, and visual dis- 
 tractors. He says: "I placed a musical box which played 
 softly in his vicinity", or "I applied a tolerably strong induc- 
 tion current to the left arm' ' , while in other series the reactor 
 "looked into a kaleidoscope with changing figures".* Ober- 
 steiner assures us that his observer "always endeavored to 
 abstract from the disturbing ifluences and to concentrate his 
 attention on the reaction".' A few years later. Boas, in con- 
 nection with his work on the difference limen, reported experi- 
 ments on the determination of the difference limen of bright- 
 nesses in which ' 'the attentive listening to a piece of music' ' was 
 used as a distraction. Here, then, we notice for the first time 
 
 iKiilpe: The Problem of Attention; The Monist, XIII, 1903, 44 f. 
 
 2 Sometimes the term distraction has been applied to one of the two 
 tasks which had to be attended to simultaneously. Or, the double- 
 task method has been called the method of simultaneous activities. 
 This has resulted in a good deal of confusion, since the single-task 
 method likewise involves simultaneous activities, the difference con- 
 sisting mainly in the direction of the attention. Therefore we have 
 thought it better to employ the new terms "single-task" and "double- 
 task" method as indicating the essential difference between the two, 
 instead of following KUlpe and calling them the methods of distracted 
 and distributed attention. 
 
 ' Wundt: Grundz. d. physiol. Psychol., ist ed., 1874, 745-749. 
 
 * Obersteiner : Virchow's Arch. f. pathol. Auat. and Physiol., LIX 
 (N. S. IX), 1874, 427-458. 
 
 5 Obersteiner : Brain I, 1879, 447 f . 
 
 8 Op cit., 449. 
 
THE MEASURBMBNT OF ATTENTION 487 
 
 that the observer did not attend to the original task of brightness 
 discrimination, but to the distraction, so that the former was 
 relegated to the periphery or background of consciousness/ 
 Boas' results were known to Stumpf, to whom they perhaps 
 suggested a new method of standardizing the different degrees of 
 attention, which we shall later discuss in detail under the head- 
 ing : method of graded distractors. The first criticism of sen- 
 sory distractors, as unable to reduce indirectly the attention 
 given to a prescribed task, was made by Cattell in 1886, in his 
 investigation of the influence of degrees of attention upon 
 reaction -time. He "let three metronomes beat and ring 
 rapidly"," and he found that "the attention can be more 
 thoroughly distracted if the brain is busied with some other 
 operation while the reactions are being made. A good way to 
 accomplish this is to let the subject beginning with any num- 
 ber add as rapidly as possible 17 after 17 to it".' It remains 
 doubtful even here whether in the few moments preceding and 
 including the reactions the subject's attention was not quickly 
 withdrawn from the addition-task. Miinsterberg has used 
 both the single-task* and the double-task^ method, employing 
 as a rule addition as a distractor. In some of Bliss' experi- 
 ments it was desirable to note the influence of attention upon 
 the task of tapping, and therefore "an effort was made to dis- 
 tract the attention of the person tapping" by weak sounds, 
 but without effect, while on the other hand ' 'the blowing of a 
 loud whistle was followed by a great irregularity", as hap- 
 pened also with "the mental addition of 214 and 23" and the 
 "mental multiplication of 14 by 5".° Swift reports that mus- 
 cular reaction-times were lengthened "while a metronome was 
 ticking one hundred and twenty times each minute".' He 
 also made some experiments "to find how the simple 'muscular' 
 and the 'choice' reactions would vary while the reactor's 
 attention was directed to certain kinds of work. Three tasks 
 were given : i , repeating a poem already committed to memory; 
 2, reading an English book; and 3, reading Kant's Kritik der 
 reinen Vemunft. . . . The instructions were to fix the atten- 
 tion as closely as possible on the work assigned".^ 
 
 In the years from 1895 to 1900 there appeared a relatively 
 large number of experimental articles bearing upon our prob- 
 lem. The single-task method was mainly used in the in- 
 
 1 Boas : PMger's Archiv, XXVI, 1881, 496. 
 
 » Cattell : Mind, XI, 1886, 237. 
 
 » Op. cit., 238. 
 
 * Miinsterberg: Zeits. f. Psychol., I, l8go, 104. 
 
 5 Miinsterberg: Beitrage z. exp. Psychol., IV, 1892, 200. 
 
 « Bliss: Stud. fr. the Yale Psychol., I<ab., I, 1893, 50. 
 
 ' Swift: Amer. Jour. Psychol., V, 1893, 5. « Op. dt., 17. 
 
488 GBISSLER 
 
 vestigations of Hamlin, Moyer, Birch, aud Darlington and 
 Talbot, done under Titchener's direction in the Cornell labo- 
 ratory. Hamlin,' repeating some of Miinsterberg's " experi- 
 ments for the purpose of measuring the effect of attention upon 
 sensible intensity, found that addition as the distracting task 
 could not be used effectively together with judging small dif- 
 ferences of intensity, brightness, and pressure, because "the 
 subjects usually found that it acted as a spur rather than as a 
 check to the attention. Their attention was wholly absorbed 
 in the addition for a part of the interval, but as a rule the dis- 
 traction was not continuous, and there came free instants in 
 which the stimulus flashed into consciousness with great clear- 
 ness and distinctness. Although the subjects were instructed 
 to give their attention to the addition, the fact that they must 
 give some judgment of the stimuli made them instinctively at- 
 tentive to the stimulus, whenever there was any break in the 
 continuity of the distraction.'" In Moyer' s experiments the 
 distractions to be attended to were, besides the mental addi- 
 tion of three-place figures, "writing the words of a sentence in 
 reverse order, writing the letters of a word backward, trans- 
 lation of simple sentences into a foreign language and writ- 
 ing the words and letters in a reverse order, and discrimina- 
 tion of odors."* Birch ° continued and extended the study of 
 odors as distractors, while Darlington and Talbot used musi. 
 cal "phrases of nine notes each, played upon the piano. Al 
 were written in J time and in the key of C. Five different 
 octaves were used, the notes ranging from the c-' to the c*". 
 Important work on attention has also been done by de Sanc- 
 tis, who in some of his experiments aimed at a measurement 
 of the intensity of concentrated and distributed attention. He 
 required the observer "to occupy himself with a (tapping) 
 movement which had to be made simultaneously with the 
 beats of a metronome and under total concentration of atten- 
 tion. . . . While the observer was thus busy, certain distrac- 
 tions were induced in him, being made progressively stronger 
 and more numerous; at the same time he was required to 
 close an electric current whenever the distraction intruded 
 into the focal consciousness.'" In order to study the varia- 
 tions of attention under the influence of different conditions, 
 Binet and Henri * have used as a distraction two metro- 
 
 1 Hamlin: Am. Jour. Psychol., VIII, 1896, 1-66. 
 
 2 Miinsterberg: Psychol. Rev., I, 1894, 39-44. 
 'Hamlin: op. cit., 49. 
 
 * Moyer: Amer. Jour. Psychol., VIII, 1896, 411. 
 5 Birch: Amer. Jour. Psychol., IX, 1897, 45-55. 
 ^Darlington and Talbot: Amer. Jour. Psychol., IX, 333. 
 'Sante de Sanctis: Zeits. f. Psychol., XVII, 1898, 209. 
 8 Binet et Henri: L'annfie psychol., Ill, 1897, 237. 
 
THB MEASUREMENT OF ATTENTION 489 
 
 nomes of slighly different rates, "par exemple 50 et 60 a la 
 minute," with the attention on the original task, while Tou- 
 louse and Vaschide * employed odors for the same purpose. 
 
 After 1900 the interest in the problem of measuring the at- 
 tention seems to have somewhat relaxed. The experimental 
 work was, however, continued by Wirth and Peters. The 
 latter attempted to find a measure of attention in the rise and 
 fall of the stimulus limen, and used the single-task method in 
 order to eliminate the Einstellung of the attention to a limi- 
 nal stimulus. " Einstellung to \hs. liminal stimulus," he says, 
 ' 'is lacking in the experiments in which the observer had to 
 concentrate upon one of the following four contents: reading, 
 mental arithmetic, reading Hungarian, pulling the spring in a 
 spring balance. ... In these experiments not only is the 
 Einstellung lacking, but we have also a new phase, the con- 
 centration upon a different content from the expected liminal 
 impression."^ There remains only to be mentioned Ebbing - 
 haus' recommendation of using "geeignete Hauptleistungen 
 und geeignete Stbrungen"' in order to obtain a more exact 
 measurement of the attention. 
 
 The double-task method seems to have been much less 
 frequently employed in psychological experiments. Perhaps 
 the first use of it was made by Loeb* in 1886. His aim was to 
 find out whether muscular activity could be used as a measure 
 of mental activity. One of his two tasks was to give ' 'the 
 maximum pressure which the flexor muscles of the hand can 
 exercise upon a dynamometer, ' ' while the observer was read- 
 ing or making multiplications of different degrees of difficulty. 
 His work was continued and extended several years later by 
 J. C. Welch, who correlated the muscular strength of the 
 hand with many different mental activities, ° for instance, vis- 
 ual and ^uditory perception and the registration of rhythmical 
 pendulum movements, mental adding and multiplication, read- 
 ing, writing, and repeating poetry. Miinsterberg has fre- 
 quently and for various purposes used adding simultaneously 
 with other kinds of observations, but he has also employed, 
 especially together with large arm-movements, ° such tasks as 
 reading, counting letters, and fixating dots. Smith investi- 
 gated "the relation of attention to our power of associating 
 and recollecting objects presented to our consciousness", and 
 
 1 Toulouse et Vaschide: Compt. rend, de see. de bid., XI, 1899, 
 
 965- 
 ^W. Peters: Arch. f. d. ges. Psychol., VIII, 1906, 406. 
 ^Ebbinghaus: Grundz. d. Psychol., I, 2d ed., 1905, 623. 
 *Loeb: Arch. f. d. ges. Phys., XXXIX, 1886, 592. 
 ^Welch: Amer. Jour, of Phys., I, 1898, 283-306. 
 ^Miinsterberg: Beitr. i.. exper. Psychol., IV, 1892, 200. 
 
 JOURNAI, — 2 
 
490 GBISSLER 
 
 found^ among other results that a "distinct but not too com- 
 plicated device for securing what one may call mental dis- 
 traction (and what he elsewhere better terms "variation of 
 attention") consisted in requiring the subject, while learning 
 the letters on the card (for the sake of memorizing them) to 
 perform a simple sum in addition, viz., that involved in repeat- 
 ing the sum 2, 4, 6, 8, . . . or rarely, when the former series 
 tended to become too mechanically easy, the series 3, 6, 9, 
 ... In order that an effective control over the behavior of 
 the subject might be secured, he was required farther to re- 
 peat the numbers aloud and to make each step in the addition 
 coincide with the stroke of a metronome beating at the rate of 
 60-70 strokes per minute."^ In Drew's experiments the ob- 
 server had to make reactions "on five and ten-fingered key- 
 boards" together with reciting the "multiplication table up to 
 12X12, . . . reading text (i) silently, (2) aloud, naming let- 
 ters (i) in direct order in text, (2) in reverse order, (3) on 
 drum, seen one at a time. . . . Reaction made by finger key 
 and with mouth key; to be made in shortest possible time."' 
 As to the purpose of this series of experiments, it is interest- 
 ing to note that Drew says: "its object was to arrange a 
 series of tasks of increasing degrees of complexity which 
 should make ever greater demands on the mind, until the at- 
 tention should pass from a fully concentrated to a completely 
 distracted state."* 
 
 Perhaps the most frequent use of the double-task method 
 has been made by Binet and Henri. For instance, they re- 
 quired an observer to press rhythmically a rubber bulb con- 
 nected with a Marey-tambour tracing on a kymograph. At the 
 same time he was asked to perform some difficult mental task 
 such as adding, reciting, or reading aloud. ° In another case, 
 Henri" required his observers to read and write simultaneously 
 two different pieces of prose. In this way Binet and Henri 
 tried to find an answer to the question "how can we know 
 whether a person gives much or little attention to a certain 
 task"' Quite recently the double- task method has received a 
 very strong impulse from McDougall, who has designed a new 
 apparatus for this special purpose. In his description of its 
 possibilities he says: "in studying the concurrent or inter- 
 current performance of two different mental operations, the 
 
 iW. G. Smith: Mind, N. S. IV, 1895, 50. 
 
 '^Op. cit., 50. 
 
 2 Drew: Amer. Jour. Psychol., VII, 1895, 533. 
 
 ^ Op. cit., 533. 
 
 ^A. Binet: Rev. philos., XXIX, 1890, 138-155. 
 
 "Henri: L'ann^e psychol., Ill, 1897, 232-278. 
 
 ' Henri: op. cit., VI, 1900, 250. 
 
THE MEASUREMENT OF ATTENTION 49 1 
 
 subject is first set to mark the spots while the cylinder rotates 
 at such a rate that he easily puts his black mark on each red 
 spot. The rate of rotation is then increased step by step until 
 a maximal effort of attention alone enables the subject to 
 mark each spot. He may then be set any other task that does 
 not involve the use of his eyes or of his right hand, /?.^., the ad- 
 dition of figures dictated to him, the counting of irregular taps, 
 the discrimination of two points touched on the skin, the per- 
 formance with the fingers of the left hand of simple or com- 
 pound reactions to auditory signals."^ Unfortunately, Mc- 
 Dougall himself has not as yet published any results obtained 
 with this instrument, although his "Preliminary Communica- 
 tion" appeared in 1905. Finally, the double-task method 
 reached the climax of its development in some of Wirth's ex- 
 periments which we have discussed at length in a previous 
 paper."" The observer had, for instance, to fixate a given point, 
 to distribute his attention in various ways over the whole 
 field of vision or over parts of it, and at the same time to ex- 
 pect and react upon a brightness change liable to appear in 
 any part of the visual field. Thus the double-task method has, 
 after all, been applied to a sufficiently large number of mental 
 and physical activities to allow a fair comparison with the 
 more frequent single-task method. 
 
 However, a true estimate of the relative value of the two 
 methods based solely upon the results obtained with them is 
 hardly possible, because in both cases the results depend very 
 largely upon the effectiveness of the distraction material em: 
 ployed. Aside from this, the two methods are probably about 
 equally well suited to serve the general purpose of varying 
 the attention to a given task. If either, the single-task 
 method has the advantage, because it is more like our every- 
 day experience, where as a rule one set of mental processes 
 occupies for some time the focus of consciousness, while others 
 are forced into the background. Some authors, it is true, 
 think that the power of distributing attention is a sign of 
 greater intelligence,' but to take up this question would lead 
 us too far from our problem. 
 
 We return, then, to a criticism of the efiectiveness of the dis- 
 traction-material. One reason for the ineffectiveness of a given 
 task, whether attended to alone or together with another task, 
 
 ^McDougall: Brit. Jour. Psychol., I, 1904, 438. 
 
 ^Geissler: A Critique of Professor Wirth's Methods of Measure- 
 ment of Attention. Amer.Jour. Psychol., XX, 1909, 120-130. 
 
 'S. de Sanctis, op. cit., 213; "dass das Vermogen, die Aufmerksam- 
 keit zu verteilen, in der Psychologie eine hohere Bedeutung hat als 
 das, sie zu fixieren." 
 
492 GEISSLER 
 
 is a lack of perfect continuity. It allows brief moments of 
 interruption, during which the other task may rise to the focus, 
 and thus be performed about as well as if it were done bj' 
 itself. This holds especially in the case of sense-discriminations 
 and reactions, which can be accomplished almost instanta- 
 neously. Another defect is that frequent repetition of the 
 same task leads to habituation, that is, it involves less and less 
 consciousness, and of course less and less attention ; thus more 
 attention can be given to the other task. Easy mental arith- 
 metic and easy prose reading as well as many muscular activi- 
 ties are especially subject to this criticism. Finally, the 
 complaint has sometimes been made that the distractions do 
 not affect all normal subjects in exactly the same way. This 
 is perhaps the main fault with odorous stimuli, although other- 
 wise they appear to have more distracting power than any other 
 source of distraction examined. Thus far it seems to have 
 been impossible to find a task which is not open to one or other 
 of these three criticisms. As Kiilpe says : " it is clear that, if 
 we are to induce a determinate degree of consciousness, we 
 must know, first, the mode of distribution of the total energy 
 of attention over the separate conscious contents included 
 within its range, and secondly the distractive value of the 
 process which we have selected for our purpose. Now no gen- 
 eral rule can be laid down under either heading ; the results 
 that have been obtained hold only in particular cases".* 
 
 The fact then that the means of varying the attention are 
 themselves but little reliable is one of the reasons why several 
 of the attempts at standardising the degrees of attention have 
 thus far failed to give satisfactory results. However, there 
 are many other reasons, which it will be best to discuss in con- 
 nection with the various methods that have been proposed for, 
 or employed in, the measurement of the attention. For the 
 sake of systematic treatment we have divided all the methods 
 into six groups. The first five may be classed together as 
 methods of expression, analogous to the physiological methods 
 employed in the investigation of the affective processes. 
 According to these five methods the degrees of attention may 
 express themselves in changes of i, peripheral vision ; 2, mus- 
 cular strength ; 3, liminal and differential sensitivity; 4, reac- 
 tion-time ; 5, accuracy of work. The sixth method may be 
 likened to the method of impression. By a series of graded 
 distractors different degrees of attention are to be induced in 
 the observer, and he is afterwards to report which degree he 
 experienced. We shall now discuss these methods in the 
 order thus presented. 
 
 1 Kiilpe: The Monist, XIII, 1903, 45. 
 
THE MEASUREMENT OP ATTENTION 493 
 
 I. The fact that Peripheral Vision is changed and narrowed! 
 under concentrated attention seems to have first been menJ| 
 tioned by Stanley Hall, in 1883. He says: " if we concentrate 
 attention upon an image at the centre of the field of vision, 
 its peripheral tracts seem to grow dark, as indeed does the 
 centre itself with some observers, when the attention is fixed 
 on a point in indirect vision"^ Experimental evidence for 
 this fact was submitted by Janet, who "concluded from it that 
 the nature of the point fixated by the subject and the degree 
 of attention she paid to it played a part in the extent of the 
 visual field". ^ To quote him in full: "it is for this that we 
 undertook a series of researches conducted as follows: At the 
 centre of the apparatus, on the central point, we fasten a piece 
 of paper on which, according to the case, some sentences or 
 numbers are written very small. We place the subject in the 
 proper position to measure her visual field ; we close one of her 
 eyes and request her not only to look at the centre but also to 
 read the paper or mentally to make up a sum with the numbers 
 written on the paper. When her attention is well fixed on the 
 work, ... we draw the stick on which the small white object 
 is fastened over the perimeter from the external side of the eye 
 that is being examined, moving from the periphery to the centre. 
 We stop a few moments at the point which we know to be 
 within the subject's field, 40°, for example. . . . We succeed 
 thus in determining the subject's visual field during the fixed- 
 ness of her attention. . . . With many normal men these con- 
 ditions do not modify the visual field ; with some they diminish 
 it from 5° to even 10°; but with hysjtericals, and in general 
 with patients whose attention is modified, this process brings 
 with it surprising contraction".' Although Janet did not find 
 a uniform contraction of the visual field with normal observers, 
 he nevertheless thinks "that the phenomena of attention do 
 not depend solely on movements (he probably has Ribot and 
 perhaps also Mtinsterberg in mind), as has been said, for the 
 modifications of the visual field cannot be connected with 
 movements".* Instead, he would rather relate these contrac- 
 tions to "the power of personal perception", i. e., the limited 
 range of attention. "In short", he continues, "it would per- 
 haps be possible to establish here the effort of attention in order 
 to synthetize phenomena". ° His facts were confirmed, as he 
 himself points out, by S6glas, who had "established them 
 independently, not aware of this first communication' ' . ° 
 
 I S. Hall : Mind, VIII, 1883, 180. 
 
 '^ P. Janet: The Mental State of Hystericals. Tr. by Corson, 1901, 71. 
 
 3 Op. cit., 71-73. 
 
 ' Op. cit., 73. 
 
 6 Op. cit., 73. * Op. cit., 71. 
 
494 GEISSLBR 
 
 However, Wtrth's recent experiments* do not seem to cor- 
 roborate them. Wirth, it is true, did not investigate the extent 
 of the visual field under various degrees of attention, but tried 
 to establish changes in the difference limen in the peripheral 
 regions of the field; nevertheless, it seems reasonable to ex- 
 pect a close relation between a higher limen and a narrower 
 field. Unless, then, Janet's results are verified by other ex- 
 periments repeating his conditions, there is no reason for 
 assuming that degrees of attention can be expressed by the 
 range of peripheral vision. Even if such a correlation could 
 be established, it would still, although valuable in itself, be 
 limited entirely to visual attention, and new methods would 
 have to be found for other departments of attention. 
 
 II. Muscular Strength exerted on a dynamometer is sup- 
 posed to express degrees of attention according to J. Loeb and 
 his pupil J. C. Welch. They assumed, in the first place, that, 
 "if P be the maximum pressure with simultaneous mental 
 work, the more attention the mental work requires the larger 
 P-p becomes' ' , and in the second place, that "if . . . f is 
 the expression of the attention in the case that the whole 
 available attention is given to muscular work, the expression 
 7^ represents the constant of attention for one kind of mental 
 work".' In other words, "the maximum static innervation of 
 the hand is arbitrarily selected here as a measure of other 
 voluntary activities simultaneously attempted", so that "the 
 determination of the constant of attention of various kinds of 
 mental work affords a method of comparing the degree of con- 
 centration of attention required in different kinds of activity".' 
 Thus the whole investigation depends upon the correctness of 
 the assumption that "^ is the expression of the attention", 
 lyoeb must certainly have known that static muscular pressure 
 or "the maximum static innervation" is not simply a function 
 of the attention, not even mainly a function of the attention. 
 Does he, then, assume that the other factors involved, espe- 
 cially the physiological factors, remain constant and are, 
 therefore, negligible? At least he fails to take them into 
 consideration, while he makes no analysis of the whole situa- 
 tion which, from the nature of the task, must have been very 
 complicated. His argument seems to rest upon the fallacy 
 that attention is analogous or even equivalent to physical en- 
 ergy and may, therefore, be added to or subtracted from bodily 
 strength and thus calculated in terms of kilogram-seconds. 
 Another objection to I<oeb's earlier work was raised by Mtin- 
 
 1 Wirth: Psychol. Stud., II, 1906, 30-88. 
 
 2 Welch: Amer. Jour, of Physiol., I, 1898, 283. 
 « Op. cit., 288. 
 
THE MEASUREMENT OF ATTENTION 495 
 
 sterberg ^ in 1888, but Welch makes no reply to it, as she 
 systematically disregards the psychological aspect and literature 
 of her problem. 
 
 III. Sensitivity, Liminal and Differential, has been proposed 
 and used as a measure of attention by several psychologists. 
 The original suggestion probably came from Fechner's wor^ 
 upon liminal intensities. The first experiments were made, 
 upon Kraepelin's proposal, by A. Bertels in 1889 with the 
 purpose of measuring not so much attention as distraction. 
 Bertels used liminal brightnesses. However, his stimuli were 
 not given simultaneously, but at determinate intervals of vari- 
 able length, lasting from several seconds to fractions of a 
 second. Besides, his assumption that ' 'a stimulus, in order to 
 reach a liminal value, must be the greater the less concentrated 
 is the attention to it",° was not at all based upon experimental 
 evidence, but simply taken for granted. Kraepelin himself 
 did not seem satisfied with the results, for in 1896 he wrote: 
 "the degree of concentration of the attention can be meas- 
 ured by determining the just noticeable stimulus intensity. 
 However, such measurements, except where quite rough dis- 
 turbances are employed, are so difficult to make and depend 
 upon so many other presuppositions that, in spite of their great 
 theoretical interest, for practical purposes they would hardly 
 come into consideration".' In recent years the liminal inten- 
 sity has been used by Wirth and Peters. The latter employed 
 as stimuli ( i ) a liminal noise made by a shot falling through 
 a variable distance upon a hard surface; (2) a liminal change 
 in the illumination of a white cardboard produced by increasing 
 or decreasing, in discontinuous steps, the resistance in an 
 electric current feeding a light of 35 c. p.; and (3) a liminal 
 pressure upon the nail of the middle finger. "The judgments 
 of the observer were: noticed (+), perhaps noticed (+?), 
 doubtful (?) , not noticed ( — ) . The observer had also to indi- 
 cate after every experiment whether, in his subjective estimate, 
 he had followed the instructions well, fairly well, or badly".* 
 In some of the experiments, "the attention had to be concen- 
 trated as keenly as possible upon the expected impression' ' , in 
 order to "obtain normal liminal values". In another series 
 the attention was concentrated upon a different activity, as 
 reading, arithmetic, or dynamometer tensions. Finally, in a 
 third test, the observer had to attend to whatever liminal stim- 
 ulus might be given, without expecting any particular one. 
 
 1 Miinsterberg : Die Willenshandlung, 1888, 160 f. 
 2 A. Bertels: Versuche iiber die Ablenkung der Aufmerksamkeit, 
 Dorpat Dissertation, 1889, 9. 
 ^Kraepelin: Psychol. Arb., I, 1896, 58. 
 * Peters : Arch. f. d. ges. Psych., VIII, 403. 
 
496 GBISSLBR 
 
 Thus the Einslellung to any one of the three possible stimuli 
 was practically eliminated. According to Peters these three 
 series represent three different degrees of attention to the 
 liminal stimulus, concentration with Einstellung, concentration 
 without Einstellung, and distraction. He finds a uniform rise 
 of the limen above the normal value in the experiments with- 
 out Einstellung, while in the distraction series the results show 
 great individual differences {cf. his Table 8, p. 416), though 
 as far as the distractions were effective, there is a tendency of 
 the limen to rise even higher than in the experiments without 
 Einstellung. Of great interest to us is also the correlation of 
 the results with the subjective estimate of the observer's atten- 
 tion to the distracting task. Peters, however, did not place 
 great confidence in these statements, because he had found 
 "long before that the criterion for the judgment of what is 
 usually called attention is not (as, for instance, Miinsterberg 
 seems to believe) uniform at all, but is variable with different 
 individuals, that one person judges the 'intensity' of his con- 
 centration by the strain sensations or by the intensity of the 
 impulses of his will, another by the clearness and distinctness 
 with which he has apprehended a conscious content, a third 
 one finally by the affective components, his interest, etc." ^ 
 Besides, Peters thinks that the estimates on different days can- 
 not be compared. This would be true, of course, if one and 
 the same observer varied on different days between the three 
 possible ways of estimation ; otherwise Peters' objection is not 
 necessarily valid. If it were true, then the same criticism 
 might be brought against judgments of intensity made on 
 different days. Since Peters knew of the variety of means of 
 estimating the attention, he might, if it had become necessary 
 for his purpose, have overcome the diflSculty by requiring his 
 observers to use only one prescribed criterion. As it is, the 
 most important addition which his investigation makes to the 
 solution of our particular problem consists in the furnishing 
 of experimental evidence for the fact that attention without 
 Einstellung is not maximally intense. 
 
 IV. Reaction- Time was first used to express degrees of 
 attention by Obersteiner. He assumed that the "retardation 
 of the reaction . . . stands in inverse proportion to the 
 intensity of attention".'' The results show that with different 
 people the same distractions have different effects. In this 
 they agree with Wundt's and Buccola's results,' who also 
 found that under distraction the time of reaction lengthened. 
 
 1 Op. cit., 428. 
 
 2 Obersteiner : Brain I, 1879, 444. 
 
 ' Buccola : Rivista di filos. scientif. I, 19. 
 
THE MEASUREMENT OF ATTENTION 497 
 
 While, on the one hand, Buccola, according to Stanley Hall,^ 
 thinks this to be "due to distraction or defective powers of 
 voluntary attention", Wundt, on the other hand, is very cau- 
 tious in the interpretation of the retardation. In the first 
 edition of the Grundz. d. physiol. Psychologie he says, indeed : 
 ' 'as long as a disturbing secondary impression is present, so 
 long it is impossible to raise the concentration of attention to 
 its highest degree".^ In later editions, however, this passage 
 is omitted. Instead we read, for instance, in the third edition: 
 "the innervation, which exists in every concentration of 
 attention, probably varies with different sense-impressions, 
 perhaps it arises in different localities of the centre of 
 apperception".' Even this is entirely rewritten in the fifth 
 edition, and the whole fact is simply referred to as "eine Ver- 
 langerung der Apperceptionsdauer des Reizes' ' . ' Nothing is 
 said here about an intensified concentration of attention or 
 apperception. Similarly careful in his interpretation of length- 
 ened reaction under distraction is Swift, who found a small but 
 uniform increase when the reaction was made while the atten- 
 tion was concentrated on some other task. He wrote: "I 
 willingly admit that it is impossible to determine how far the 
 instructions are observed in such a case".* He even believed 
 that ' 'there are no means by which the amount of attention 
 given to the reaction or to the other task can be measured". 
 Cattell," who distinguished three introspective degrees of atten- 
 tion, concentrated, normal, and distracted attention, found that 
 there was no corresponding lengthening of the reaction-time 
 with less attention. Sharp, on the other hand, thinks it pos- 
 sible that "an individual who gave a quick reaction-time, 
 owing to great power of attention, might, by having his atten- 
 tion artificially distracted, lengthen his reaction time until it 
 corresponded to that of other individuals who normally gave 
 long reaction times from small power of attention. Given a 
 constant correspondence between length of reaction-time and 
 degree of concentration of the attention, and the rapidity of 
 reaction of an individual might be taken as an index of his 
 power of attention".' Nevertheless she doubts the range of 
 its practical application, while to us the basal fault seems to 
 lie in the fact that "the constant correspondence" of which she 
 speaks as given is not given, but is a very dubious matter, 
 
 1 Hall : Mind, VIII, 1883, 177. 
 
 3 Wundt: op. cit., ist ed., 1874, 749. 
 
 'C/ cit., 3rd ed., II, 1887, 293. 
 
 * Op. cit., 5th ed., Ill, 1903. 
 
 ^ Swift: Amer.Jour. of Psychol., V, 1895, ^1- 
 
 sCattell: Mind, XI, 1886, 242. 
 
 'Sharp: Amer. Jour, of Psychol., X, 1899, 16. 
 
498 GEISStER 
 
 which has not yet been established by experimental methods. 
 Binet and Henri also made use of the reaction experiment, 
 not, however, for the sake of its retarded character, but for its 
 irregularities as expressed by the mean variation. Quite re- 
 cently Wirth and Kastner ' have combined their perimetrical 
 brightness-experiments with the reaction to the recognition of 
 the brightness change, but their results are much too irregular 
 to give any indication as to the different degrees of attention 
 under which the task was done. Thus the final outcome of 
 the reaction experiments, as used for the measurement of atten- 
 tion, has been on the whole negative ; it has been impossible 
 to establish a positive correlation between high degrees of 
 attention and short reactions, and between lower degrees and 
 correspondingly lengthened reactions. 
 
 V. '^\\& Degree of Precision, as the last method to be men- 
 tioned in this connection, seeks to find an expression of the 
 degree of attention in the quality and quantity of work per- 
 formed in a given time. At first sight this method seems to 
 be the most natural and promising of all, but even here we 
 find disagreement as to its efi&ciency. Some psychologists 
 have assumed that, in general, changes in the quality or quan- 
 tity of the work done are due directly to changes in the amount 
 of the attention given to the task. Peters ^ calls this Bourdon's 
 method. It has been recommended, for instance, by Jodl : "it 
 is obvious that attention cannot be measured directly. How- 
 ever, its degrees may be determined indirectly by reference to 
 its effects. These effects consist, in every task which does not 
 have the character of an involuntary movement, in the avoid- 
 ance of errors and mistakes and in the quickness of the exe- 
 cution".' Lipps, as we have pointed out, likens attention to 
 a "psychisch^ Kraft" which, like any kind of power, may be 
 measured by the "Hohe der Leistungen".* In a similar way 
 Stern holds that "the quality and quantity of a task performed 
 under normal conditions could be used as a measure of the 
 attentional energy' ' - ^ This method has been used, for example, 
 by Sharp : "the degree of attention habitually exercised by an 
 individual was measured by the quickness and accuracy with 
 which a certain given task was performed". ° Hence her seven 
 observers were classified accordingly, the fastest and most ac- 
 curate as giving the best attention, the slowest and least 
 accurate as giving the least attention to the same task. Vari- 
 
 * Wirth and Kastner: Psychol. Stud., IV, 1908, 139-200. 
 ^Peters: op. cit., 390. 
 
 ' P. Jodl : Ivehrbuch d. Psychologie II, 1903, 84-85. 
 
 * Th. Lipps : Leitfaden d. Psychologie, 1903, 34. 
 
 ^W. Stern: Cber Psychol, d. individ. Differenzen, 1900, 84. 
 « Sharp: Amer.Jour. of Psychol. X, 1899, 28. 
 
THB MEASUREMBNT OF ATTENTION 499 
 
 ations of the attention of a single individual at different times 
 could, of course, be detected and treated in a similar way. A 
 modification of this method was first suggested by Friedrich 
 in 1883. He said: "it is tempting to fix definitely the some- 
 what unsettled concept of attention by making it proportional 
 to the measure of precision, i. e., to the reciprocal value of the 
 mean error, so that a small mean error corresponds to a high 
 degree of attention, and conversely, a large mean error to a 
 low attention, . . . since the characteristic phase of in- 
 attention consists rather in the large variations of the physio- 
 logical times within the same experimental series' ' . Although 
 his results showed "a very good agreement between this theo- 
 retical consideration and experience' ' , nevertheless he thought 
 that ' 'the identification of the attention with the measure of 
 precision is in most cases open to objections. Only in cases 
 of the simplest mental processes, which besides should be as 
 homogeneous as possible and little subject to practice, may one 
 assume that the mean error is mainly dependent upon the 
 degree of attention. With more complex mental processes, 
 however, not only does practice play a considerable part in the 
 fluctuations of the individual experiments, but in certain cases 
 the cause of great variations may lie in the mental processes 
 themselves' ' . -^ The experiments of Friedrich were made under 
 Wundt's direction just before the publication of the second 
 edition of the Grundz. d. physiol. Psychologic, where Wundt 
 himself made the first change in his interpretation of the re- 
 tarded reaction-time, so that it is probably safe to assume that 
 Wundt agreed with Friedrich's contention. Nevertheless, the 
 mean variation has been frequently proposed and employed as 
 a measure of attention. H. Griffing, for example, says: 
 "although we cannot assume that the average number of letters 
 seen by an observer measures his powers of attention, the mean 
 variation from the average of the numbers seen in the different 
 experiments is presumably due, principally at least, to varia- 
 tions in the attention".^ A wide application of the mean 
 variation as a "dynamometer of the attention" is found in 
 Oehrn's experiments. He gives,' for instance, the following 
 m. V. values : for writing 2.6%, reading 3.4%, reading a foreign 
 language (Latin, Livy) 3.3%, counting letters 4.2%, arith- 
 metic 4.6%, counting letters in groups of three 5%, memoriz- 
 ing i2-place numerals 14.7%, and learning series of nonsense- 
 syllables 27.4%, of the average value in the particular task. 
 However, this work is criticised by Henri thus: "we do not 
 
 1 M- Friedrich : Philos. Stud., I, 1883, 73 f. 
 ^Griffing: Amer.Jour. of Psychol., VII, 1895, 235. 
 8 A. Oehrn: Psychol. Arb., I, 1896, 121 f. 
 
500 GBISSLER 
 
 believe that these figures express exactly the part due to 
 attention in any one of the above cases, since other factors, 
 like fatigue and practice, are not eliminated and certainly in- 
 fluence also the mean variation".^ On the other hand, van 
 Biervliet^ maintained that "the attention can be measured by 
 the consistency of the results, in fact, by the mean variation", 
 aind he applied this method in his experiments on visual acuity. 
 Binet, reviewing and criticising this article, makes the follow- 
 ing peculiar comment: "il y a dans I'article de M. Biervliet 
 une remarque tres ingenieuse, qui est a conserver. II dit que 
 I'attention se marque dans la variation moyenne d'une experi- 
 ence ; c'est tout a fait juste, dans beaucoup de cas. On le 
 savait, mais personne jusqu'ici ne I'avait dit avec cette net- 
 tete".° Peters, whose work we have already discussed, found 
 a close correlation between the magnitude of the m. v. in de- 
 terminations of the liminal sensitivity and the degrees of 
 attention under which he had established the limen. Never- 
 theless, even so he had to assume that practice had ' 'a slight 
 but distinct" * influence upon the results. Thus the mean 
 variation can at best afibrd only a very rough measure of at- 
 tention. Perhaps its greatest usefulness would consist in 
 furnishing a convenient check upon other methods and results, 
 especially since it is not bound up with any particular appara- 
 tus or procedure of experimenting, as is the case with all the 
 other methods of expression.^ 
 
 This review of experimental work brings to light the im- 
 portant fact that the expression methods have almost entirely 
 neglected the factor of introspection. The main reason for this 
 has probably been correctly stated by Peters; the investigators 
 have not been able to agree upon a common characteristic feature 
 of the state of attention. Some of them have emphasized 
 the motor factor, some the affective components, some the 
 strain sensations, but with the exception of Wirth none have 
 used the clearness factor. This oversight is the more sur- 
 prising as most of the theoretical discussions of the degree of 
 
 ^Henri: L'ann^e psychol.. Ill, 1896, 244. 
 
 2 Van Biervliet: Journal de psychol., I, 1904, 231. 
 
 'A. Binet: L'ann^e psychol. XI, 1905, 71. 
 
 * Peters: op. cit., 426. 
 
 ^ In addition to the five methods of expression thus far indicated 
 we briefly mention the following method reported by Bliss, op. cit., 
 pp. 53 i. "During the course of my experiments Dr. Scripture 
 suggested that the accuracy with which a person could steadily 
 point to a given spot would be a measure of the amount of attention 
 he could direct toward the work. la accordance with that suggestion 
 the apparatus shown in Fig. 15 was arranged to measure this accuracy. 
 . . . (The result was that) within certain limits the movements of 
 the point increased or decreased inversely with the amount of atten- 
 tion given to the work". 
 
THB MEASUREMENT OF ATTENTION 501 
 
 attention have explicitly or implicitly referred to clearness and, 
 as a matter of fact, have agreed upon it as one of the main 
 characteristics of attention, while as to the other characteristics 
 they heartily disagree. 
 
 VI. The only systematic attempt to employ introspection as 
 an aid toward measuring the attention has been made by the 
 method of Graded Distractors , as we may briefly call it, which 
 is therefore the only representative of the impression-method 
 in attention. By means of a series of stimuli which should 
 make greater and greater impression upon the observer's 
 mind, he would be induced to pay less and less attention t<}. 
 another simultaneous task. Stumpf, who considers it one of 
 the two problems of a quantitative study of attention "to find 
 means by which degrees of attention may be differentiated",^ 
 thinks that such means can be found, and gives as an example 
 ' 'the simultaneous perception of other simple phenomena which 
 are capable of a progressive multiplication. The amount of 
 this multiplication furnishes, of course, not a direct but an in- 
 direct measure of attention. Attention itself is not, but the 
 means of varying it are, measurable' ' . "^ Stumpf refers to Wundt 
 and Boas as having already used this principle in their experi- 
 mental work. Later it was applied by Drew in certain ex- 
 periments. Their object was "to arrange a series of tasks of 
 increasing degrees of complexity which should make ever 
 greater demands on the mind, until the attention should pass 
 from a fully concentrated to a completely distracted state".' 
 
 Although he accumulated bulky introspective records, Drew 
 does not state in how far he succeeded with his series of dis- 
 tractions. After Drew, Stumpf's principle was made the 
 leading problem of a series of minor studies in the Cornell 
 laboratory, carried out during the years 1895-97. 1*o quote 
 Titchener: "Suppose, again, that we have arranged a series of 
 distracting stimuli, homogeneous in kind but graded in com- 
 plexity, such that we are able to reduce the observer's per- 
 centage of right cases from 100 to 95, 90, 85 . . . according 
 to the distraction employed. It is necessary that the action of 
 the distractors be constant ; and it is necessary that they be 
 of the same kind, and therefore exert an influence which difiers 
 only, and difiers measurably, in degree. Having secured these 
 conditions, we should let the observer decide whether the clear- 
 ness of conscious contents was distinguishably different under a 
 5 per cent, and a lo per cent, distraction, or under a 5 per 
 cent, and a 15 per cent, distraction, or again under an 80 per 
 cent, and an 85 per cent, distraction, and so on, all through the 
 
 J Stumpf : Tonpsychologie, I, 1883, 73. 
 
 ' Op. cit; 75. ° Drew: Amer. Jout. of Psychol., VII, 1895, 533. 
 
502 GEISSLER 
 
 series. We should thus finally obtain a scale of noticeably 
 different clearnesses paralled by a scale of measured amounts 
 of distracting stimulus' ' . ^ However, the experiments have not 
 yet led to the discovery of an entirely satisfactory series of 
 graded distractors . It was found "that odors are admirably con- 
 stant distracting material, — if only they could be measured"!^ 
 Thus the impression method has had little more success than 
 the expression methods, although psychologically it seems to 
 be the sounder,' because it attempts to parallel experimental 
 results with introspective data. The reason for its failure thus 
 far is not inherent in its nature, but lies rather in the difficulty 
 of finding satisfactory material, namely, effective distractions 
 which are at the same time objective, homogeneous, and quan- 
 titatively measurable. 
 
 We have thus completed our review of the experimental 
 work upon the problem of measuring the attention and may 
 conclude, on the one hand, that, although much effort has 
 been spent upon it, we are still far from a satisfactory solution. 
 On the other hand, we are now better able to judge which 
 paths would seem most promising for future experiments and 
 which should be abandoned. We are also convinced that 
 Pillsbury's suggestions on "« priori grounds"^ are not at all 
 hopeful. We miss in his chapter on "The Methods of Measur- 
 ing Attention" the reference to the clearness factor, whose im- 
 portance he emphasized in his introductory chapters, 
 
 PART II 
 
 ^ An Attempt at a New Measurement of Attention in 
 Terms of Clearness Values 
 
 The general aim of the following investigation was to re- 
 new the search for an exact method of measuring the atten- 
 tion by advancing simultaneously along old lines which the 
 previous review of the literature had shown to be most promis- 
 ing, and along new paths which suggested themselves in the 
 course of the experiments. Previous investigations had simply 
 assumed that the variations of attention are directly paralleled 
 by certain changes, whether in reaction-times, or in muscular 
 pressure, or in quickness and accuracy of work, etc., etc., but 
 they had not attempted to verify this assumption by reference 
 to experimental introspection, and thus their psychological 
 task was only half completed. 
 
 The purpose of our own work was to determine whether 
 
 1 Titchener : I,ectures on the El. Psychol, of Feel, and Att., 1908, 277. 
 
 2 Op. cit., 278. 
 
 '^Cf. Ebbinghaus: Grundz. d. Psych., I, 1905, 621 f. 
 * PiUsbury : Attention, 1908, 92. 
 
THE MEASURBMENT OF ATTENTION 503 
 
 and in how far the assumed parallelism of the expression- 
 methods rested on an experiential basis. This test might be 
 applied in several different ways. For instance, one might 
 argue that, if the assumption made by the expression-methods 
 holds true, then it must be possible to find a direct parallel 
 between any two ways of expressing degrees of attention, or 
 rather, conversely: if two kinds of operations, which seem to 
 depend mostly upon a strong concentration, give a positive 
 correlation between different degrees of perfection, then the 
 assumption holds true that their different degrees of perfection 
 express different degrees of the attention with which they 
 were executed. Or, one might argue thus: it would seem un- 
 reasonable to hold that objective variations in the work ac- 
 complished should indicate subjective variations in the obser- 
 ver's attention, if his careful introspection could not reveal 
 such variations. 
 
 The former principle was applied in a first group of experi- 
 ments, in which we compared the accuracy and quantity of 
 work performed with the muscular rigidity or motor inhibi- 
 tion which is said to become the more pronounced the more 
 attention is concentrated. Ebbinghaus, for example, has 
 pointed this out in the following passage: "the same inhibi- 
 tory relation which exists between purely mental states and 
 processes, occurs also between them and bodily movements. 
 In medium degrees of intensity both may quite well take place 
 simultaneously. As soon, however, as on the one side the 
 energy of activity is considerably increased, so soon will the 
 other side suffer a decrease."^ This fact had never been 
 tested experimentally, at least for the purpose of measuring at- 
 tention by it, unless one should include here I^oeb's and 
 Welch's dynamometer experiments. In order to get a graphic 
 record of muscular rigidity the following arrangement was 
 made. The observer's left arm was screened from his sight, 
 and he was required to grasp with the hidden hand a handle 
 fixed in an arm rest. The observer had no knowledge of the 
 existence and purpose of the further details of the apparatus. 
 The handle was in reality a plethysmograph, which was con- 
 nected by glass and rubber tubing with a Marey tambour 
 writing on a kymograph. The slightest movement of any of 
 the five fingers, or any tightening or loosening of the grip, 
 caused an appreciable excursion of the pointer, which increased 
 with the extent of the movement. 
 
 The work which, after several trials with various materials, 
 seemed best suited for an objective, quantitative gradation as 
 to quickness and accuracy, was mental arithmetic in the form 
 
 1 Ebbinghaus: Grundz. d. Psychol., I, 1905, 597. 
 
504 
 
 GEISSI,ER 
 
 of continuous adding. We arranged 4 diflFerent series of 30 
 figures eacli, selected so as to constitute 4 difiFerent degrees of 
 difficulty. The easiest series contained all figures from 3 to 13, 
 except ID, each figure occurring 3 times. The second degree 
 of difficulty included the following figures: 23, 24, 25, 26, 27, 
 33-37. 43-47. and 53-57, each figure occurring twice. In 
 the next series the following figures were used: 63-67 once, 
 73-77 and 83-87 twice, and 93-97 once. Finally the most 
 difficult series contained: 128, 129, 131-135,221-227, 314-319, 
 408, 409, 411-413, and 503-507. By using different starting 
 points from day to day, practice or memory of the re- 
 sults was entirely eliminated, while the actual series did not 
 need to be changed. The results are given in Table I. The 
 following distractions were used: an electrically driven tuning 
 fork of 512 vibrations, an electric bell, a metronome beating 
 
 Tabi,e I 
 
 Difficulty I. 
 
 Difficulty 2. 
 
 Difficulty 3. 
 
 Difficulty 4. 
 
 
 To: 25.36,47. 
 
 To: 65, 66, 68, 
 
 To: 29, 31, 34, 
 
 Tj: 18,22,29, 
 
 
 58, or 64 
 
 72, 74, or 77 
 
 35. or 38 
 
 32, or 38 
 
 
 Add: 
 
 Add: 
 
 Add: 
 
 Add: 
 
 
 9 
 
 27 
 
 63 ■ 
 
 314 
 
 
 II 
 
 34 
 
 74 
 
 409 
 
 
 6 
 
 45 
 
 85 
 
 135 
 
 
 7 
 
 53 
 
 97 
 
 315 
 
 
 12 
 
 37 
 
 Zl 
 
 503 
 
 
 13 
 
 55 
 
 86 
 
 411 
 
 
 8 
 
 46 
 
 77 
 
 226 
 
 
 5 ■ 
 
 36 
 
 96 
 
 224 
 
 
 3 
 
 24 
 
 87 
 
 128 
 
 
 9 
 
 57 
 
 73 
 
 504 
 
 
 4 
 
 26 
 
 67 
 
 412 
 
 
 II 
 
 47 
 
 76 
 
 316 
 
 
 7 
 
 35 
 
 64 
 
 227 
 
 
 12 
 
 25 
 
 83 
 
 222 
 
 
 8 
 
 43 
 
 75 
 
 129 
 
 
 3 
 
 37 
 
 84 
 
 505 
 
 
 13 
 
 54 
 
 73 
 
 221 
 
 
 4 
 
 35 
 
 83 
 
 413 
 
 
 5 
 
 47 
 
 95 
 
 319 
 
 
 II 
 
 33 
 
 V- 
 
 225 
 
 
 6 
 
 45 
 
 87 
 
 131 
 
 
 9 
 
 56 
 
 86 
 
 506 
 
 
 12 
 
 34 
 
 65 
 
 134 
 
 
 7 
 
 23 
 
 93 
 
 317 
 
 
 3 
 
 44 
 
 76 
 
 133 
 
 
 6 
 
 '^5 
 
 94 
 
 507 
 
 
 13 
 
 36 
 
 85 
 
 318 
 
 
 5 
 
 46 
 
 77 
 
 223 
 
 
 8 
 
 33 
 
 66 
 
 408 
 
 
 4 
 
 44 
 
 84 
 
 132 
 
 
THE MEASURBMBNT OP ATTENTION 505 
 
 at 3 different rates, 60, 100, and 150 strokes per minute, and 
 a flicker moving across the whole field of vision at 4 different 
 rates of speed. 
 
 All experiments were done in a large dark room. Directly in 
 front of the observer, at a distance of 2 m, stood a large screen, 
 1.20 x2.40 m, covering practically the whole field of vision 
 when the eyes fixated the central point. It was illuminated by 
 an electric light above and somewhat behind the observer's 
 chair. In front of this light a rotating spoke, made of a strip 
 of cardboard, cast its flickering shadow over the visual field. 
 A rectangular hole of 3 x 8 cm was cut in the centre of the 
 screen, and through it the figures to be added were exposed. 
 The figures of a series were pasted on a long strip of cardboard 
 sliding in a groove directly behind the screen. To the upper 
 end of the strip a string could be attached, which ran over a 
 pulley and carried a counter weight. Another string, tied to 
 the lower end of the strip, led to a spool with a handle, in front 
 of the observer's seat. By turning the handle he wound the 
 string around the spool, thus pulling down the strip and ex- 
 posing the figures one after another at any desired speed. 
 
 As soon as the experimenter had announced the initial figure 
 the observer began to add to it the figures exposed, calling out 
 each individual sum until all 30 figures of a series had been 
 added. The observer was instructed to concentrate all his at- 
 tention upon the adding, and to neglect as far as possible any 
 distraction, purposive or accidental. At the same time the 
 experimenter behind the screen checked up each individual 
 sum, marking every mistake, and also obtaining a graphic 
 record of the time of each individual addition. This was done 
 in the following way: one of the two ends of a pair of com- 
 passes carried a pencil with which the accuracy was recorded, 
 while around the other end a copper wire was wound which 
 was connected with an electromagnetic pointer writing on the 
 kymograph. Next to the strip of paper upon which the 
 accuracy of the additions was recorded, a long strip of sheet 
 copper was fastened, connected by a wire with the other pole 
 of the magnet. As soon as the observer pronounced a sum the 
 experimenter touched the copper strip with the wired compass 
 end; thus the electric circuit was closed, and the pointer raised 
 by the magnet made a mark on the kymograph. A Jacquet 
 clock furnished the time line by writing seconds. Thus for 
 every experiment three graphic records were obtained; the 
 movements of the observer's left hand, the time line, and 
 the duration of each single addition as well as that of the whole 
 experiment. The kymograph with the writing pointers was 
 enclosed in a soundless box, a lever being so placed that the 
 experimenter, by raising or lowering it with his left hand, 
 
 JOURNAI,— 3 
 
506 GBISSLSR 
 
 could easily and quickly start and stop the movement of the 
 kymograph. After this was done, his left hand was free to 
 manipulate any of the distractions to be employed during the 
 experiment, while his right hand held the compass ready to 
 check up the results. 
 
 After every experiment the observer was required to write 
 put or to dictate answers to the following introspective ques- 
 tion ary: 
 
 (i) What was your general affective mood during the experiment? 
 
 (2) Did it undergo any change, and if so of -what kind and during 
 which part of the experiment? 
 
 (3) Where and how intensive was your attention? 
 
 (4) Did it undergo any noticeable changes? If so, of what kind 
 were they and when did they occur ? 
 
 (5) Did you notice any internal sensations during the experiment? 
 If so, of what kind were they and when did they occur? 
 
 (6) Did they distract you and to what extent? 
 
 (7) Did yon notice any external stimuli? Of what kind were they? 
 
 (8) Can you remember in which order they occurred? 
 
 (9) Did they distract yon and to what extent? 
 
 10) Can yon add any other introspective observations not included 
 in the previous questions? 
 
 It should be noted that questions 3 and 4 did not suggest 
 any criterion for distinguishing different degrees of attention. 
 The observers judged the intensity of their concentration of 
 attention in terms chosen entirely by themselves. After some 
 experience and practice in estimating their attention they de- 
 veloped, independently of one another and without sugges- 
 tions from the experimenter, a scale of 4 or 5 different degrees 
 of concentration. This fact was utilized in the second half of 
 the experiments belonging to this group, by the adoption of 
 the following arbitrary scale of degrees: very good or excellent, 
 good, fair, bad, very bad; without, however, any definition of 
 the nature of these degrees. 
 
 The experiments of the first group, including all prelimi- 
 nary and practice series, lasted from November, 1 907, till March, 
 1908, when they were discontinued because they promised 
 nothing new or of further significance. The observers in this 
 as well as in the next group of experiments were Dr. W. H. 
 Pyle (P), Dr. T. Nakashima (N), and Mr. R. A. Tsanoff 
 (T). All three were graduate students in the psychological 
 department, and had had considerable practice and experience in 
 introspection. They observed regularly three times a week. 
 Four experiments were made in the hour, and a rest of from 5 to 
 10 minutes was allowed between the second and the third, 
 during which the observer could leave the dark room. In 
 every hour, each one of the four degrees of difiiculty occurred 
 once. Their temporal order was so arranged that every 
 difi&culty occurred an equal number of times in the first, 
 
THE MEASUREMENT OF ATTENTION 507 
 
 second, third or fourth place. A long preliminary practice 
 was required from the observer, and also from the experimen- 
 ter. The addition tasks used in these experiments were 
 slightly different in arrangement, but not in difficulty, from 
 those employed in the later series. The regular series began 
 in January, 1908, each observer making at first 24 experiments 
 without any distractions. In the following 300 experiments, 
 about 100 for each observer, the distractions occurred with the 
 following frequency: the tuning fork, the electric bell, and 
 each one of the four rates of flicker occurred three times, while 
 each one of the three metronome rates occurred twice with 
 every one of the four degrees of difficulty. The distractions 
 were so distributed as to let each kind occur only once in an 
 hour, and as often in the first place as in the other places. 
 However, they did not persist during the whole experiment, 
 but were used only during the first twenty additions, while 
 the last ten additions were always performed in the normal way. 
 
 The results showed that, in spite of all precautions, a slow 
 and gradual effect of practice and habituation made itself felt, 
 especially during the first half of this group of experiments. 
 The practice both reduced the length of time for the different 
 degrees of difficulty and decreased the number of errors. The 
 habituation manifested itself in a more and more pronounced 
 disregard of the distractions, so that the first twenty addi- 
 tions were done as well as the last ten, and in some cases bet- 
 ter, because the sudden interruption of the tone or flicker 
 might confuse the observer for a few moments. 
 
 Considering the results as regards the degree of difficulty, 
 we notice a uniform increase in the average length of time 
 from the easiest to the most difficult series. Table II gives 
 the results in seconds for the first 24 normal experiments, for 
 all 96 experiments with distractions, and for the last 48 ex- 
 periments with distractions. The item Limits shows the 
 quickest and the slowest experiment that occurred with each 
 degree of difficulty. 
 
 The overlapping between neighboring degrees of difficulty is 
 due to practice, which brought it about that toward the end 
 a more difficult task could be done as fast as an easier task had 
 been done at the beginning. It is interesting to notice tha^, 
 without exception, the more difficult additions reached the 
 maximum of practice sooner than the easier additions. The 
 shortest times for the former occurred after about six weeks of 
 regular tri-weekly experimenting, while for the latter the 
 shortest times occurred much later. The only difference of 
 effectiveness between the different distractions was that the ex- 
 periments with the flicker gave as a rule much shorter times 
 and fewer errors than those with auditory distractions. Among 
 
5o8 
 
 GEISSLER 
 Tablb II 
 
 
 Difficulty I. 
 
 Difficulty 2. 
 
 Difficulty 3. 
 
 Difficulty 4. 
 
 Obs. 
 
 Av. 
 
 Limits. 
 
 Av. 
 
 Limits. 
 
 Av. 
 
 Limits. 
 
 Av. 
 
 Limits. 
 
 P 
 
 N 
 T 
 
 90 
 
 108 
 
 75 
 
 75- 96 
 90-132 
 66-84 
 
 no 
 
 137 
 117 
 
 105-116 
 121-159 
 100-135 
 
 146 
 185 
 154 
 
 134-160 
 
 149-233 
 136-160 
 
 163 
 209 
 180 
 
 136-214 
 189-221 
 166-201 
 
 P 
 
 N 
 T 
 
 73 
 85 
 57 
 
 57-98 
 7X-117 
 41- 96 
 
 92 
 
 no 
 103 
 
 72-111 
 
 87-143 
 81-131 
 
 109 
 
 145 
 116 
 
 89-146 
 
 116-188 
 
 84-145 
 
 128 
 171 
 152 
 
 100-167 
 
 139-237 
 114-190 
 
 P 
 
 N 
 T 
 
 71 
 80 
 
 49 
 
 62- 84 
 69-98 
 41- 61 
 
 85 
 102 
 100 
 
 72- 94 
 87-119 
 81-128 
 
 100 
 141 
 103 
 
 89-115 
 
 116-153 
 
 84-123 
 
 ii6 
 I6S 
 138 
 
 105-137 
 139-191 
 
 IJ14-173 
 
 the latter themselves no differentiation as to effectiveness 
 could be observed. 
 
 In comparing the observers' estimates of attentive concen- 
 tration with the calculated quickness and accuracy of their 
 work, we shall refer only to the last 48 experiments, because 
 here a uniform scale of degrees was used and the influence of 
 practice was minimal as to improvement both in rate and in 
 accuracy. In fact the number of errors was so small as to be 
 entirely negligible. In order to discover whether there was 
 any correlation between the observer's estimate of attention 
 and the calculated time-values of his work, the latter were 
 divided into 5 groups : quickest rate, fairly quick, average 
 rate, fairly slow, and slowest rate. Because the number of 
 experiments available from each observer was small, the results 
 were combined for the purpose of calculation, after we had 
 assured ourselves that the individual results were sufficiently 
 similar to allow such a procedure. The 5 introspective degrees 
 of attention were then paired with the 5 rates of work to give 
 the following Correlation-Table : 
 
 Tabi,e III 
 Degrees of Attention 
 
 
 
 Best Good 
 
 Av. Bad Least 
 
 Total No. 
 
 Means 
 
 Rate 
 
 of 
 
 work 
 
 Quickest 
 
 Quick 
 
 Average 
 
 Slow 
 
 Slowest 
 
 24 1 
 
 13 17 
 
 22 17 
 
 8 6 
 
 — 4 
 
 6 1 — 
 i6 I — 
 
 3 4 — 
 
 25 
 37 
 56 
 21 
 
 5 
 
 1.04 
 1.86 
 
 1-93 
 2.14 
 2.20 
 
 
 Total No. 
 Means 
 
 67 45 
 2.21 2.91 
 
 26 6 
 2.96 3.5 
 
 144 
 
 r 
 
 = .60 
 
THE MEASUREMENT OF ATTENTION 509 
 
 The correlation coefiBcient was calculated according to the 
 familiar Bravais- Pearson formula r^= , ^^ ■ ■ which resulted 
 
 in the equation ■^^^~^=-(^o. This is a surprisingly high 
 
 value, considering the fact that the observers were not told in 
 what terms to define degrees of attention. They were, how- 
 ever, frequently warned not to judge these degrees on the basis 
 of an estimation of the accuracy or quickness of the work, 
 and as a rule they faithfully followed this instruction. The 
 introspections do not, however, throw much light upon the 
 methods of estimating attention used by the various observers. 
 Occasional hints indicate a few general facts : in the case of 
 P, the estimation was mostly based upon the ease with which 
 the different sums occurred in consciousness. N usually relied 
 upon his memory of the force with which the distraction im- 
 pressed him ; while T used as a basis the presence or absence 
 of motor concomitants or "general nervousness". However, 
 these criteria do not seem to have been used consistently or 
 uniformly, nor were the observers always aware of their own 
 methods. This fact tends to make the close correlation be- 
 tween the subjectively estimated degrees of attention and the 
 objectively determined quantity of work the more remarkable, 
 and should lead to greater confidence among psychologists in 
 the observer's ability to make reliable estimates of quantitative 
 aspects of his own consciousness. Our results are entirely 
 different from those obtained by Peters, and our later experi- 
 ments show that the trustworthiness of introspective estimation 
 of degrees of attention can be still further increased by the 
 adoption of a uniform principle of differentiating such degrees. 
 On the other hand, the plethysmographic curves, that were 
 to have shown the muscular rigidity of the left hand, gave 
 absolutely negative results. In most cases they were perfectly 
 smooth, while with one observer they gave sometimes distinct 
 pulse records. Out of 128 experiments with P, only 6 showed 
 brief irregularities, with N this number was 17, with T 12; 
 and most of them occurred in the very earliest experiments. 
 Hence it was impossible to discover any kind of correlation 
 between the state of bodily rigidity and mental work. In this 
 respect, then, the application of our principle of comparing 
 two expression-methods failed, mainly because of an unfortu- 
 nate choice of the methods to be compared. The principle 
 itself is not affected by these negative results, and awaits 
 further test. 
 
 Since the results of the first group of experiments had inci- 
 dentally shown how well even unsystematic introspection is 
 
5IO GBISSLER 
 
 able to reveal subjective variations of attention, they offered a 
 sufficient inducement for continuing these experiments as an 
 explicit application of the second principle mentioned above. 
 In other words, our problem shifted to this : How closely are 
 the introspective estimates of attention correlated with quanti- 
 tative changes of an objectively measurable character? In 
 order to obtain systematic uniformity in the results, it was ad- 
 visable to prescribe to the observer the terms in which he was 
 to estimate the degrees of his attention. We followed Titch- 
 ener's suggestion in asking the observers to base their sub- 
 jective estimates of attention upon differences in the clearness 
 of the mental processes attended to. In general, the procedure 
 in this second group of experiments remained the same as in 
 the first group. A few slight changes in method and material 
 will be mentioned later, but the plethysmographic handle was 
 not removed, chiefly because we wished to keep the external 
 conditions of the experiments as constant as possible. 
 
 In order that the observers might become perfectly familiar 
 with the introspective differences of clearness, they completed 
 a large number of preliminary experiments, which were later 
 repeated with those who observed in the third group of 
 experiments. Besides, several undergraduate students, who 
 at the time were members of the introductory laboratory course, 
 were asked to take part in these preliminary experiments, 
 mainly in order to see how easily they learned to distinguish 
 between differences of clearness and differences of intensity. 
 
 The method in the preliminaries was in general this: two 
 metronomes were set going at different rates, about loo and 
 1 20 strokes per minute, and the observer, who sat with his 
 back toward them, was asked to count the number of sounds 
 between coincident strokes. Then the same stimuli were re- 
 peated, but the observer was required to perform some mental 
 operation, for example, spelling or reciting in a foreign 
 language, adding or subtracting successively the same number, 
 naming rapidly the states with their capitals along the Atlantic 
 coast, etc. Each experiment should not last longer than a 
 minute or two. After several paired experiences of this kind, 
 the observer is not only able to compare the differences between 
 the clearness of the metronome strokes in the first and in the 
 second case, but is also ready to assert that the strokes were not 
 always equally clear or obscure. As soon as this stage is 
 reached, the experiments are repeated with the same or similar 
 material and the observers are encouraged to estimate, first the 
 larger, later the finer, differences in clearness. Some observers 
 will easily construct for themselves a scale of five or more steps 
 from very clear to very obscure, while others will prefer to use 
 rough differences in terms of per cent. Nearly all of our 
 
THE MEASURKMBNT OF ATTENTION 511 
 
 observers, both experienced and inexperienced, found it rela- 
 tively easy to apply some such gradation of clearnesses. In 
 all these preliminaries the differences between the focal and 
 the peripheral processes were rather large. The efforts to in- 
 duce clearnesses which lay nearer the middle of the arbitrary 
 scale were most successful when the observer was requested to 
 hum to himself a more or less familiar melody and gently beat the 
 time to it with hand or foot, while either one or both of the two 
 metronomes were striking their own rhythms. Another effec- 
 tive method was to spell long phrases, or to add or subtract 
 series of figures, to the rhythm of a metronome. After several 
 days of practice the more experienced observers were able to 
 give a fairly complete analysis of their consciousness during 
 an experiment, and to estimate without difficulty the clearness 
 or obscurity of every mental content experienced by them. It 
 seemed, at least according to some of the observers, that the 
 comparison of simultaneous processes as to their degree of clear- 
 ness facilitated especially the task of analyzing a complete con- 
 sciousness even in entirely different experiments. 
 
 After eight or ten hours of practice the observers P, N, and 
 T had worked out for themselves, with suggestions as to a 
 uniform terminology from the experimenter, the following scale 
 of nine degrees of clearness: 
 
 I. About 90-100%= maximally clear or excellent attention 
 
 very clear ' ' very good ' ' 
 
 clear " good " 
 
 fairly clear " fairly good " 
 
 fair " fair 
 
 fairly vague " fairly poor " 
 
 vague " poor " 
 
 very vague " very poor " 
 
 20%= absolutely obscure " no " 
 
 I^ater, during the experiments in the dark room, all the 
 observers added, of their own accord and independently of one 
 another, an intermediate step between the first and the second 
 degree, while T made one also between the second and the 
 third degree. 
 
 The experiments of the second group were resumed in the 
 dark room about the middle of April. The distractions which 
 had been used in the first group were again employed, but 
 now in various combinations, in order to find out whether 
 this would make them more effective for inducing different de- 
 grees of attention. In the four simplest cases two auditory 
 distractors were combined in the following way: 
 
 I. two metronomes of different rates, loo and 120 per 
 minute. 
 
 2. ' 
 
 ' 80- 90%= 
 
 3- ' 
 
 ' 70- 80%= 
 
 4- 
 
 ' 60- 70%= 
 
 5- ' 
 
 ' 50- 60%= 
 
 6. ' 
 
 ' 40- 50%= 
 
 7- ' 
 
 ' 30- 40%= 
 
 8. ' 
 
 ' 20- 30%= 
 
 9- ' 
 
 0- 20%= 
 
512 
 
 GBISSLER 
 
 2. one metronome and the electric bell, 
 
 3. one metronome and the electric tuning fork, 
 
 4. the bell and the fork. 
 
 In the next four combinations the visual distraction caused 
 by a medium rate of flicker was added to these pairs. Finally, 
 three simultaneous distractors were given in this way: 
 
 1. two metronomes of difierent rates, the bell, and the 
 fork, 
 
 2. two metronomes of different rates, the bell, and the 
 flicker, 
 
 3. two metronomes of different rates, the fork, and the 
 flicker. 
 
 The addition tasks were continued with one degree of diffi- 
 culty only, but the figures 43-47, 53-57. 63-67, and 73-77 
 were given in the four different orders of Table IV. The 
 number of single additions in one experiment was reduced 
 from 30 to 20, so that the last ten (without any distractions) 
 were omitted. Instead of this, during every hour one of the 
 four experiments was completed normally, that is without any 
 distractors, in order to be used as a standard of comparison. 
 The normal series, as well as the series with the various com- 
 binations of distractions, were so distributed over the whole 
 group of experiments as to occur with equal frequency in the 
 
 Table IV 
 
 To: 113, 115, 
 
 To: 221, 225, 
 
 To: 332. 335. 
 
 To: 444. 445, 
 
 or 117 
 
 or 229 
 
 or 338 
 
 or ,146 
 
 Add: 
 
 Add: 
 
 Add: 
 
 Add: 
 
 56 
 
 47 
 
 46 
 
 73 
 
 45 
 
 63 
 
 64 
 
 55 
 
 67 
 
 56 
 
 73 
 
 64 
 
 74 
 
 44 
 
 57 
 
 47 
 
 54 
 
 73 
 
 43 
 
 74 
 
 73 
 
 55 
 
 66 
 
 54 
 
 63 
 
 76 
 
 55 
 
 65 
 
 44 
 
 66 
 
 77 
 
 56 
 
 55 
 
 77 
 
 6s 
 
 75 
 
 77 
 
 46 
 
 54 
 
 46 
 
 64 
 
 65 
 
 47 
 
 63 
 
 46 
 
 54 
 
 74 
 
 45 
 
 53 
 
 45 
 
 63 
 
 53 
 
 65 
 
 67 
 
 53 
 
 67 
 
 76 
 
 53 
 
 48 
 
 76 
 
 57 
 
 ?5 
 
 75 
 
 57 
 
 47 
 
 64 
 
 67 
 
 77 
 
 43 
 
 74 
 
 56 
 
 44 
 
 75 
 
 57 
 
 45 
 
 43 
 
 66 
 
 43 
 
 76 
 
 66 
 
 first, second, third, or fourth place of an hour's work. After 
 3 days' preliminary work with normal series only, each ob- 
 
THE MBASUREMBNT OF ATTENTION 513 
 
 server completed 60 experiments, of which 44 were done with 
 distractions, so that each distractor combination occurred four 
 times. 
 
 The questions for introspection were also somewhat modi- 
 fied, in accordance with the uniform standard of diflferentiating 
 degrees of attention, and now ran as follows: 
 
 (i) Making the relative clearness of the adding process the crite- 
 rion for degrees of attention, and calling the absolute maximal clear- 
 ness which you can experience under most favorable conditions 100% 
 or excellent attention, what percentage of attention or what degree of 
 clearness prevailed, during the experiment, for the adding process ? 
 
 (2) Was it constant, or intermittent, or changing to more or less? 
 
 (3) On the same basis of distinction, what degree of clearness or 
 attention would you ascribe to other mental processes that occurred 
 during the experiment? 
 
 (4) Were they uniform, or variable, and how? 
 
 (5) Were you distracted by anything else? If so, by what and 
 how much? 
 
 (6) Was the distraction continuous or intermittent? 
 
 (7) What affective processes or moods dominated your conscious- 
 ness during the experiment? 
 
 (8) Can you add any other appropriate remarks? 
 
 Besides, the observers were again expressly cautioned not 
 to estimate their attention by the apparent temporal rate or 
 the seeming correctness of their calculations. The general 
 instruction was to concentrate attention upon the adding- 
 process and to neglect as much as possible any intentional or 
 accidental distractions. The large number of introspective 
 data obtained from these, as well as from the other observers 
 in later experiments, furnish interesting facts concerning the 
 steadiness of attention and the problem of the two levels, 
 which will be discussed later. 
 
 The results of this second group of experiments showed 
 plainly that even the most complex combinations of distractors, 
 after a few days' work, were unable to induce great variations 
 of attention. Instead, toward the end of the whole group, 
 most of the normal series were actually performed at a slower 
 rate than the distraction series. Throughout the whole course 
 of this group of experiments a slight and gradual practice 
 effect was noticeable which, in the case of P and N, mani- 
 fested itself in both a shortening of the time and in smaller 
 deviations from the general average. In the case of T, how- 
 ever, a rather sudden improvement occurred about the middle 
 of the whole course, while his variations from the average re- 
 mained as great as in the beginning. 
 
 Although the distractions were not as effective as expected, 
 yet the attention to the adding-process was not the same in all 
 experiments, but varied between the first and the fifth degree ; 
 most frequently it was judged "very good", in a few cases it 
 
514 GEISSLHR 
 
 sank to the fourth degree, once it was estimated by N at ' 'fair' ' , 
 and once at "fairly poor"; but all these lower degrees occurred 
 during the first few days, before the distractions had lost much 
 of their effectiveness. In the case of P attention was most 
 frequently judged at the highest degree, and only once or 
 twice below "very good". This agrees perfectly with the 
 quality of his work, the rate of which is much faster and the 
 quality more uniform than that of the other two observers. 
 Table V allows a detailed comparison of the subjective esti- 
 mates of attention in terms of clearness with the average time 
 (in seconds) of the experiments at these levels. With all three 
 observers there is a perfect correlation between their best at- 
 tention and their shortest time, and between correspondingly 
 lesser degrees and longer times. The fractions given under 
 the item Error have for their numerator the number of ex- 
 periments belonging to each rubric, and for their denominator 
 the number of errors occurring in these experiments. Wherever 
 the m. V. is given as o, only one experiment occurred under 
 that rubric. 
 
 Tablk V 
 Degrees of Attention 
 
 O Expts. Av 
 
 First Betw. 1-2 Second Betw. 2-3 Third 
 . MV E Av. MV E Av. MV E Av. MV E Av. MV B 
 
 4 and 5 
 Av. B 
 
 P Total 54 
 N No. 67 
 T Expts. — 
 
 4-2 if 57 3-7 \\ 
 7-7 1 69 3.7 \\ 
 60 9.6 if 
 
 63 5-5 H 
 71 7.8 if 
 
 64 9-5 II 
 
 66 4- 
 79 9-5 ¥ 
 
 78 8.6 1 
 87 5-0 1 
 
 95 % 
 
 P Normal 59 
 N 67 
 T Expts. — 
 
 3-2 1 54 3-3 f 
 7.7 1 71 2.6 1 
 63 13. 1 
 
 63 2.0 f 
 78 5-7 % 
 69 6.8 f 
 
 66 i 
 76 4-2 4 
 
 69 1 
 
 -_-_ 
 
 P Distr'n 53 
 
 N — 
 T Expts. — 
 
 4.4 ^ 58 3-4 \\ 
 
 66 3.4 1 
 
 55 9-4 if 
 
 62 5.0 J/ 
 69 7-7 If 
 64 9-5 \\ 
 
 
 
 
 83 11 % 
 
 79 8.5 ? 
 87 5.0 % 
 
 95 t 
 
 In the whole Table there is but one case in which a lower 
 degree of clearness shows a shorter average time than the next 
 higher degree. This is due to the fact that all 4 experiments 
 of the lower degree occurred within the last two weeks of work, 
 while 6 of the 8 experiments in the higher degree came in the 
 first three weeks, so that in this particular instance a cumu- 
 lative practice effect obscures the correlation. On the whole, 
 however, the shortening of the times due to practice was so 
 gradual and constant that the correlation still holds, with but 
 few exceptions in the case of N, if the whole period of experi- 
 menting is divided into 3 equal parts of 5 days' work each. 
 We omit the table, because it is very like Table V. In a 
 
THB MBASUREMENT OF ATTENTION 5 15 
 
 similar way, we also fractionated the results with reference to 
 the number of distractions used, and found that the correlation 
 in this case holds with practically the same uniformity. The 
 fact that the difference in the averages of two successive col- 
 umns is only a few seconds is not surprising, if one considers 
 that the difference in clearness itself is at best only very slight. 
 On the other hand, the great regularity of these results and 
 the general uniformity of the correlation, in spite of the very 
 small differences in degrees of attention, are the best evidence 
 for the general validity of our conclusion that the degrees of 
 attention may be very accurately estimated by introspection, if 
 they are distinguished in terms of the clearness of the mental 
 process attended to. 
 
 Because the experiments of the second group involved only 
 the higher degrees of attention, and since they were not numer- 
 ous enough for certain degrees, our next problem was to find 
 a method that should induce all possible variations of attention 
 to a given task. At the same time certain other requirements 
 had to be fulfilled. In the first place, the efficiency of the 
 work performed must depend as exclusively as possible upon 
 the degree of attention given to it and as little as possible upon 
 such factors as practice, fatigue, mood, etc. In the second 
 place, the performance should never become entirely automatic 
 or habitual, no matter how frequently it was repeated. Thirdly, 
 it must require absolutely continuous attention, so that a 
 momentary lapse should at once manifest itself in a momentary 
 reduction of the quality of work. lyastly, the execution of 
 the work must easily submit itself to a scale of quantitative 
 gradations. 
 
 The kind of work which seemed to fulfill these conditions best 
 of all was a modification of W. McDougall's method of mark- 
 ing irregularly arranged and step-fashion exposed colored dots 
 or circles.^ The exposure was made by means of the memory 
 apparatus furnished by Spindler and Hoyer. The largest of 
 the 5 aluminum cylinders belonging to it has a circumference 
 of 48cm, is 7.7cm high, and gives 24 successive exposures. 
 We fitted a large neutral-gray cardboard around the black 
 metal exposure frame, in order to screen from the observer the 
 whole apparatus and the experimenter. An electric light of 
 16 b.p. was fixed just above the exposure slit in the metal 
 frame, so as to give a constant illumination and to do away 
 with the shadow which by daylight the observer's hand and 
 pencil cast upon the visual field. Another small screen pro- 
 tected the observer's eye from the direct light of the bulb. An 
 
 ' McDougall: Brit. Jonrn. of Psychol., I, 1904, 438. 
 
5l6 GBISSLER 
 
 armrest was adjusted to the table in order to support the right 
 arm from elbow to wrist, and to bring the hand exactly in front 
 of the exposure slit. 
 
 The rings or circles to be marked were the common mathe- 
 matical sign for the circle, approximately 3 mm in diameter 
 (or lo-point of printers' size), as the sample shows. They 
 
 
 
 
 
 
 
 
 
 
 were printed in red ink upon strips of white paper which could 
 be pasted tightly around the cylinder. The circles were ar- 
 ranged in 5 different irregular patterns which could be used in 
 either direction so as to give really 10 different series. The 
 lateral or horizontal deviations were also much larger than in 
 McDougall's case, allowing for 5 different positions of the 
 circles within a space of 32 mm, thus making the distance be- 
 tween two neighboring positions 8 mm. The greatest lateral 
 deviation between two successive circles was 24 mm, or the 
 length of three positions; however, this occurred only once or 
 twice on each strip. The series thus far described will be 
 referred to as "normal patterns", in distinction from the fol- 
 lowing "complication patterns" which were also employed. 
 In one kind of series, which may be called the ' 'letter pattern" ^ 
 
 amtdvOlui 
 
 YBL0AGPFZ 
 
 LS CU0VKHM 
 
THE MBASUREMENT OF ATTENTION 517 
 
 the circles appeared in rows of capital letters ( lo-point printers' 
 size), each of the 24 rows on a strip containing 8 letters, their 
 lateral distance from each other being 8 mm. The whole strip 
 thus forms a sort of printers' pi, with the red circles mixed in. 
 In another kind of series, the "digit pattern", the numerals 2 
 to 9 were substituted for the letters, so that each numeral 
 occurred only once in a row. The figures show some 
 
 6742 ©6938 
 
 83 0579264 
 
 2) 79 ©53648 
 
 samples of each of the three patterns (with the exception that 
 the circles are not painted in red). Finally another series was 
 the "color pattern", in which little colored paper disks, of 
 the size of the circles, were pasted on the strips of paper, 
 in place of the letters or numerals. The colors used were blue, 
 yellow, red, green, black, neutral gray, blue-green, purple, 
 reddish orange, and yellowish orange. They occurred in ir- 
 regular order, the last four also with irregular frequency, but, 
 like the letters and numerals, about equally often in each of 
 the nine possible lateral positions. 
 
 The marking of the red circles, which was done with a pointed, 
 medium soft pencil, consisted in drawing a straight vertical 
 line through the centre of the circle so that its length was 
 about twice that of the diameter and extended the length of a 
 radius above and below the periphery. Through the slit, 
 2x6.5 cm, in the black metal screen in front of the cylinder, 
 one circle, or one whole row of letters, numerals or colors, with 
 a circle among them, was exposed at a time for a fraction of a 
 second, coming to a momentary rest and then disappearing 
 quickly while the next row came into sight. The speed of the 
 successive exposures could be regulated by means of adjustable 
 fans, and it was so chosen for different observers as to be the 
 fastest rate at which they could do the best possible work. 
 The two observers who took part in the final series marked the 
 24 circles in 12 seconds; the three observers in the preliminaries, 
 
5l8 GBISSI<BR . 
 
 who had less practice, varied between 15 and 22 seconds. 
 The reason for drawing a line, instead of making a dot, as in 
 McDougall's experiments, was that the former way of marking 
 secured greater continuity of attention by filling out the entire 
 interval of exposure, thus preventing momentary wanderings 
 of attention, or at least revealing them by a bad mark. 
 
 In determining the quality of the task, the first four marks 
 in each experiment were disregarded, in order to eliminate the 
 effects of the initial imperfect motor preparedness. The 20 re- 
 maining marks were graded in the following way. On the 
 basis of 100% for an absolutely perfect record,, each individual 
 perfect mark received 5%. A single error, which reduced its 
 value to 4% , consisted either in a slight variation of the length 
 or the direction of the line, or in the failure to pass through the 
 centre. If two such errors occurred together, or if the line 
 touched the periphery of the circle, the mark was graded 3%. 
 If three errors came together, or if peripheral displacement was 
 combined with a change in either length or direction, the value 
 of the mark was 2%. Finally, if the line did not touch the 
 circle at all, or if it could not be classified under any of the 
 above headings its value was 1%, while an entirely omitted 
 mark was graded zero. The sum of the 20 individual values 
 thus represents an exact measure of the accuracy of marking 
 during any given experiment. 
 
 For the purpose of inducing a greater range of degrees of 
 attention to the marking test, our experimental arrangement 
 offered numerous possibilities, because it involved three kinds 
 of variable factors. In the first place, the stimulus itself could 
 be either the normal pattern or one of the three combination 
 patterns. In the second place, the task could be either a 
 single task, namely marking the circles only, or a double task, 
 namely marking the circles while performing simultaneously 
 some mental operation like spelling, calculating, reciting, 
 memorizing, etc. (other kinds are mentioned also by McDou- 
 gall). Finally, the direction of the attention was variable ; it 
 could be shifted from the marking to the mental operation or 
 even to the accompanying letters, numerals or colors. 
 
 In the autumn of 1908 a number of preliminary experiments 
 were done, mostly for the sake of a general orientation. The 
 observers were Prof. I. M. Bentley (B), Dr. W. H. Pyle (P), 
 and Mr. T. Okabe (O), the latter a graduate student in the 
 department of psychology, with considerable experience in 
 introspection. B and O took part in 30 and 40 experiments 
 respectively, while P finished as many as 60. These prelimi- 
 naries showed that the method was adequate to our purpose 
 and that it also fulfilled the other requirements. Hence the 
 experiments were continued in February and March, 1909, with 
 
" 
 
 „ 
 
 
 tt It 11 
 t( tt tt 
 
 (( 
 
 (C 
 
 
 tt tt tt 
 
 and 
 
 
 
 
 Spelling 
 
 (( 
 
 it 
 
 (( 
 ti 
 tt 
 
 
 Spelling *' 
 
 (1 it tt 
 
 n 
 
 tt 
 
 
 tt tt (( 
 
 and 
 
 
 
 
 Reciting 
 
 tt 
 
 
 Reciting " 
 
 tt tt tt 
 
 tt 
 
 tt 
 
 
 t( tt tt 
 tt it n 
 
 £S only 
 tt 
 
 To the 
 
 tt tt 
 tt tt 
 
 Letters only- 
 Digits " 
 Colors " 
 
 THE MBASUREMENT OF ATTENTION 519 
 
 Miss H.-M. Clarke (C) and Miss A. de Vries (V) as observers. 
 Both were pursuing graduate work in psychology and had 
 several years' training in experimental work. They each com- 
 pleted 130 of the final experiments, in which the following 
 combinations of the three variables occurred : 
 
 stimulus Task Direction of Attention 
 
 1. Normal pattern Marking circles only To the Marking task only 
 
 2. Letter pattern " 
 
 3. Digit pattern " 
 
 4. Color pattern " 
 
 5. Normal pattern " 
 
 6. Letter pattern " 
 
 7. Digit pattern " 
 
 8. Color pattern " 
 
 9. Normal pattern " 
 
 10. Letter pattern " 
 
 11. Digit pattern " 
 
 12. Color pattern " 
 
 13. Letter pattern Marking circles only 
 
 14. Digit pattern " 
 
 15. Color pattern " 
 
 Other corabinations had also been tried in the preliminaries, 
 but with less success; for example, interpreting as letters cer- 
 tain tactual impressions upon the left hand. In the combina- 
 tions 13 to 15 the direction of the attention to the letters, 
 digits, or colors, was aided by requiring the observer to form 
 ajudgment, say about the most frequent or least frequent let- 
 ter, numeral, or color, or about some combination of them. 
 In the case of the combinations 5 to 12 the marking of the 
 circle was done on a lower level of consciousness ; it had, so to 
 speak, to take care of itself. 
 
 From 5 to 8 experiments could be finished in one hour. At 
 the end of each experiment the following questionary was 
 answered : 
 
 1. How much attention (in terms of clearness) was given: 
 
 (a) to the marking of the circles? 
 
 (b) to the other required task? 
 
 (c) to other sensory or ideational processes? 
 
 2. How steady was this clearness (or: within which limits did it 
 
 vary), and which degree of clearness was the most frequent 
 with regard to : 
 
 (a) the marking of the circles? 
 
 (b) the other required task? 
 
 (c) the other mental processes? 
 
 3. What affective mood prevailed during the experiment? 
 
 4. At how much is the quality of marking the circles estimated in 
 
 terms of per cent.? 
 
 5. What other relevant comments can you make upon this experi- 
 
 ment? 
 
 There was at first some difficulty in the estimation of the 
 
520 
 
 GEISSLBR 
 
 clearness of the marking. This task involves the visual per- 
 ception of the circles and their positions as well as kinses- 
 thetic sensations from the quick motor adjustments of the hand 
 and the movements of the pencil. After some practice, how- 
 ever, these factors fused into a conscious complex of homoge- 
 neous clearness which was easily estimated. Rare exceptions 
 to this were always revealed in the introspections, and if in 
 such cases the differences in clearness between the three factors 
 were only slight, their average was accepted as a final estimate, 
 while if the differences were more than two degrees of clearness, 
 the experiment was disregarded and repeated later on. In a 
 few cases, observer V felt unable to make a judgment about 
 the clearness of the marking, because her attention was too 
 intensely directed to some other task. These doubtful cases 
 were classed in the ninth degree, which represents ' 'least or 
 no attention." 
 
 The results of the marking tests agree perfectly with those 
 of the adding tests, and extend the correlation between degrees 
 of attention and quality of work to the lower levels of clear- 
 ness. Table VI shows the average value and the mean varia- 
 
 Tabi,e VI 
 
 Obs. 
 
 First 
 Av. MV 
 
 Second 
 Av. MV 
 
 Third 
 Av. MV 
 
 Degrees of Attention 
 
 Fourth 
 Av. MV 
 
 Fifth 
 Av. MV 
 
 Sixth 
 Av. MV 
 
 Seventh 
 Av. MV 
 
 8th and 9th 
 Av. MV 
 
 C 
 V 
 P 
 
 76 4.9 
 
 77 3-4 
 85 2.1 
 
 71 3.0 67 5.3 
 
 703-5 635.3 
 
 77±2.9 
 
 6r 4.5 
 SI 5-3 
 
 55 5-0 
 47 5-0 
 
 7i±6-3 
 
 49 6.0 
 43 4-0 
 
 55±8 
 
 40 6.9 
 
 41 5-6 
 5 
 
 29 7.6 
 40 5.6 
 34 ir-3 
 
 tion of the experiments made at the various clearness levels. 
 Only in the case of P the values obtained at two levels were 
 combined, because they were not very numerous, while the 
 results of B and O, although they show a similar tendency to 
 correlation, were too few to justify their tabulation. The 
 results of C and V were treated also by the Bravais-Pearson 
 formula. For this purpose, the whole range of the calculated 
 values of work, from the lowest to the highest per cent, was 
 divided into nine equal parts, which were then compared with 
 the nine degrees of clearness. The correlation-coeflBcients 
 obtained under this arrangement were .83 and .73 from C and 
 V respectively. These two figures show beyond doubt that 
 introspectively distinguishable variations of attention are very 
 closely paralleled by corresponding differences in the accuracy 
 of work performed at these different levels of attention, pro- 
 vided that the estimation of degrees of attention is done in 
 terms of clearness values, and that the work itself is as little 
 
THE MEASUREMENT OF ATTENTION 52 1 
 
 as possible influenced by anything else but changes in atten- 
 tion. 
 
 The objection might be raised that our observers based 
 their judgment of degrees of attention upon an implicit estima- 
 tion of the value of their work ; this estimation is easier, since 
 daily life offers many opportunities for practice in judging one's 
 own work and ability. If this were the case, however, the 
 correlation between the estimated and the calculated value of 
 the work should at least be just as great as, if not greater 
 than, the correlation between estimated attention and calcu- 
 lated work. But the correlation coefficients for estimated and 
 calculated work were .76 and .28 for C and V respectively, as 
 compared with .83 and .73 in the previous correlation. This 
 difference shows plainly that our observers used a criterion for 
 the judgment of degrees of attention different from that em- 
 ployed in the estimation of the quality of the work. What 
 their criteria were can be best shown by quoting from a summary 
 introspective description given by C and V at the close of the 
 entire work with the marking tests. 
 
 Observer C who, like V, preferred to estimate clearness 
 differences in terms of per cent. , employs ' 'a sort of schematic 
 visual image of 100, rather complex," but somewhat different 
 from her customary visual image of the whole number-scale. 
 She described the latter as a line going out to the right in a 
 horizontal plane, while the visual clearness scale is a line that 
 goes out straight in front of her, with its farther end slightly 
 raised. She goes on to say : 
 
 "Clear out at the other end I can see visually the 100, and that is 
 the clearest part for me of the whole thing, which is becoming very 
 shadowy toward me. The clearness degrees represent themselves to 
 me by the approach they make to the 100 mark. It is a matter of 
 clearness in memory; if the circles stand out plainly in memory 
 ( after the experiment) I put them high on the scale, if not, I put 
 them low. There seems to be a cumulative effect of the twenty-four 
 circles in any experiment. Sometimes, however, a single circle stands 
 out strongly, but this does not affect the whole estimate. ... I 
 find a great difficulty in noticing strain, because something else is so 
 much more interesting that the strains are too much in the back- 
 ground. Sometimes the strains are noticed later (after the experi- 
 ment), as e. g. in a fatigued hand, but ordinarily they are too much 
 in the background to be remembered even later." As to her judg- 
 ment of the quality of work, C says: "the ease with which I mark 
 enters sometimes into my estimates. I also had often some kind of a 
 general impression as to how many good and bad marks there were. 
 If I could not remember this I took it for granted that my work was 
 poor It was always pleasant when the work seemed easy, and un- 
 pleasant when it seemed hard. The cumulative effect of these 
 individual feelings influenced my judgment of the quality of the 
 work." 
 
 These introspections are in close agreement with her re- 
 
 JOtJRNAI,— 4 
 
522 GBISSI<BR 
 
 suits. She has fairly definite, reliable, and constant criteria 
 both for her clearness estimates and for her judgments of 
 work, and hence her correlation coefficients are remarkably 
 high, namely .83 and .76 respectively. She was never in 
 serious doubt about either of her two estimates, and hesitated 
 only rarely in cases of the very lowest degrees. In this re- 
 spect she differs widely from observer V, who in 20 out of the 
 130 experiments was unable to tell what kind of work she 
 had done, especially if her attention had not been directed to 
 the marking. V was also much more uncertain about the 
 lowest degrees of clearness, and the averages of work (given 
 in Table VI) for these lower levels are very little different 
 from one another. Her final introspections indicate that she 
 was aware of this weakness, and they also show plainly that 
 her clearness estimate could not have been influenced by an 
 evaluation of her work. She says: 
 
 "I valued the circles in terms of clearness. Sometimes I could dis- 
 tinguisli between their visual clearness and the motor clearness of 
 marking. Sometimes I was not conscious of seeing the circles at all. 
 The thing attended to was always very much clearer than the others, 
 which receded the more into the background, the higher the other 
 process was. My standard is either maximal or minimal, that is, I 
 estimate from "no attention" upward, or from "maximal attention" 
 downward. With my strongest attention there is no strain sensation, 
 but after the experiment is over I feel relaxed, from which I infer 
 that I must have been strained before, although I did not notice it, 
 because my attention was too good on the task. The scale was not 
 visualized; it is simply a matter of clearness of the object attended to. 
 The quality of work had never influenced the estimate of the atten- 
 tion. In fact the estimate of the work was mostly a guess. I never 
 remember the marking of the circles collectively. If I saw myself 
 making a bad mark I judged my work usually very poor. When my 
 attention is on some other task, I certainly do not focus my eyes 
 upon the circles, and the marking seems to do itself. It must be 
 analogous to the automatic writing of answers to questions without 
 hearing the questions." 
 
 There can be no doubt, therefore, that the objection men- 
 tioned above does not hold. 
 
 The main problem of our investigation has thus received a 
 satisfactory solution. We have shown that the assumed 
 parallelism of the expression-methods, especially as far as the 
 measure of precision is concerned, rests upon solid introspec- 
 tive evidence, if clearness is accepted as the essential charac- 
 teristic of attention. At the same time we believe that our 
 results have brought us within reach of a new and fairly defi- 
 nite method of measuring attention, for the results have shown 
 that degrees of clearness are just as accessible to introspective 
 determination as variations of the intensity of sensations. 
 This was somewhat surprising to ourselves, because we 
 doubted in the beginning the adequacy of introspection, and 
 
THE MEASUREMENT OF ATTENTION 523 
 
 expected at best only a rough parallelism between estimated 
 degrees of attention and calculated degrees of precision in 
 work. There is sufficient warrant in our results for the 
 assumption that continued practice will lead at least to a differ- 
 ential clearness limen which may be just as definite and pre- 
 cise as any other psychophysical difference limen, while on 
 the other hand it may be perhaps more difficult to establish 
 the least possible or the highest possible clearness degree for 
 any given mental process. However, the determination of a 
 difference limen for clearness would be the most important 
 step toward an exact measurement of the concentration of at- 
 tention. It would enable one, by starting with a certain 
 clearness degree of a given mental process under fixed experi- 
 mental conditions, either to increase or decrease that clearness 
 by just noticeable differences, until the maximum or the mini- 
 mum of attention to the particular process is reached. 
 
 The results of our experiments also give some indication 
 that the increase or decrease in clearness depends not so much 
 upon the nature and number of distracting stimuli as upon 
 certain other factors. The chief among these seem to be the 
 kind and complexity of the tasks involved and the instructions 
 or directions for subsequent introspection. Peters had already 
 shown that attention with Einstellung is more concentrated 
 than without it. Now the Einstellung itself depends largely 
 upon the Aufgabe or instruction. If only one task is to be 
 performed, as, for instance, in our case, the marking of the 
 circles in the normal and the combination patterns, it is neces- 
 sary before every experiment to state at least whether maximal 
 or normal attention is required, otherwise the Einstellung is 
 determined by chance and will result in a larger range of de- 
 grees of attention. If two tasks are to be performed simul- 
 taneously, for example, the marking together with the reciting, 
 the spelling, or the counting of certain combinations, then it 
 must be indicated which task is to be attended to, and to what 
 extent the other may be neglected. Such an instruction is 
 easier to follow than the requirement to give to both tasks an 
 equal amount of attention, especially if they are very different. 
 How far the nature and complexity of the tasks to be per- 
 formed determine the degree of attention given to them is 
 shown in Table VII. The vertical columns of figures headed 
 by the 9 degrees of attention indicate the frequency with which 
 the marking in different experiments was done at the various 
 levels. The individual means for each observer as well as the 
 combined means are shown in the last two columns, while the 
 column before gives the total number of experiments in each 
 case. In the first four rows of the Table the clearness of the 
 marking test when done alone is compared with the clearness 
 
524 
 
 GBISSLER 
 Tabi,E VII 
 
 
 
 
 Degrees of Attention 
 
 
 Means 
 
 No. 
 
 Nature of Bxpte. 
 
 Obs. 
 
 I 
 
 2 
 
 3 4 
 
 56 
 
 789 
 
 T. 
 
 Ind. 
 
 Comb. 
 
 I 
 
 Norm, and Comb'n-P 
 Marking Circles 
 
 C 
 V 
 
 9 
 
 13 
 
 20 
 13 
 
 9 5 
 ID 7 
 
 I 
 I 
 
 
 44 
 44 
 
 2.32 
 2.31 
 
 2-3 
 
 2 
 
 Norm, and Comb'n-P 
 Marking and Reciting 
 
 C 
 V 
 
 
 3 
 
 2 8 
 I 4 
 
 14 I 
 67 
 
 8 2 
 
 28 
 28 
 
 4.28 
 5-81 
 
 50 
 
 3 
 
 Norm, and Comb'n-P 
 Marking and Spelling 
 
 c 
 
 V 
 
 
 
 2 4 
 3 
 
 9 7 
 5 8 
 
 4 2 
 7 5 
 
 28 
 28 
 
 5-46 
 6.21 
 
 5-8 
 
 4 
 
 Combination-P 
 Attention to Comb'n 
 
 C 
 V 
 
 
 
 3 2 
 
 3 
 
 6 6 
 
 48 
 
 7 3 3 
 II 2 2 
 
 30 
 30 
 
 6.1 
 6-3 
 
 6.2 
 
 5 
 
 Normal Pattern 
 All possible Tasks 
 
 C 
 
 V 
 
 4 
 
 lO 
 
 I 
 
 98 
 3 4 
 
 63 
 S 3 
 
 I 
 3 4 
 
 40 
 40 
 
 3-4 
 3-67 
 
 3-54 
 
 6 
 
 Letter-Pattern 
 All possible Tasks 
 
 c 
 
 V 
 
 I 
 I 
 
 6 
 
 2 
 
 3 3 
 
 4 3 
 
 3 6 
 
 6 
 8 2 I 
 
 30 
 30 
 
 4-43 
 5-37 
 
 4.9 
 
 7 
 
 Digit-Pattern 
 
 All possible Tasks 
 
 c 
 
 V 
 
 2 
 2 
 
 5 
 
 2 
 
 I 4 
 4 2 
 
 9 4 
 38 
 
 121 
 
 7 I 
 
 29 
 29 
 
 4-55 
 50 
 
 4.8 
 
 8 
 
 Color-Pattern 
 All possible Tasks 
 
 c 
 
 V 
 
 2 
 
 3 
 
 r 
 
 3 3 
 3 5 
 
 56 
 56 
 
 332 
 8 2 I 
 
 30 
 30 
 
 5-0 
 5-58 
 
 5-3 
 
 of the same test when most of the attention is given to reciting, 
 to spelling, or to some combination of letters, digits, or colors. 
 The latter kinds of task require very high degrees of clearness 
 for themselves and permit only very little attention to be given 
 to the marking, namely about the 6th or 7th degree as a rule, 
 the two means being 6.1 and 6.3 for C and V respectively. 
 Spelling allows more attention to be given to the marking, and 
 reciting still more, while of course the marking alone may in- 
 volve the very highest degree of clearness. The influence of 
 the complexity of the patterns is shown in the last four divi- 
 sions of the Table. Work done with the normal pattern per- 
 mits as a rule the highest attention to the marking, the letter 
 and the digit patterns allow very much less, while the color 
 pattern allows the least attention to the marking of the four 
 kinds of patterns. However, we do not wish to place too 
 much emphasis upon these results, because the number of ex- 
 periments is not large enough for general conclusions of this 
 sort to be based confidently upon them. 
 
 Finally, another point in regard to which the large intro- 
 spective material accumulated in the course of our investigation 
 reveals some interesting evidence, is the number of simulta- 
 neous clearness levels and their relative distance from each 
 
THS MBASUREMENT OF ATTENTION 525 
 
 Other. It is frequently maintained that extreme concentration 
 to one group of mental processes is accompanied by an equally 
 extreme obscurity of other mental processes forming a general 
 background, and that less concentration is simultaneous with 
 correspondingly less obscure background processes. Pillsbury 
 expresses this by saying : "the amount of attention is practi- 
 cally constant and cannot be applied to one object without 
 affecting the clearness of others".* So far as our knowledge 
 goes, such statements have been based on casual observations 
 only, and we have, therefore, thought it advisable to present 
 the following Table VIII as giving a numerical expression to 
 the facts in question. The 9 large horizontal divisions indi- 
 cate the 9 clearness degrees which by introspective estimation 
 were assigned to the marking test. The figures in the third 
 vertical column tell how frequently with each one of the three 
 observers O, P, and V, the marking was estimated at any 
 particular level. The next 9 vertical columns of figures show 
 in how many cases other mental processes simultaneous with 
 the marking at a certain level were mentioned under a certain 
 degree of attention. Thus, for instance, in 24 experiments O 
 estimated his marking at the highest level and in the same 
 experiments assigned to other simultaneous processes in 2 cases 
 the 6th degree of clearness, in 10 the 7th degree, in 12 the 8th 
 degree, and in 6 cases the lowest or 9th degree. In some of 
 these cases an observer was able to mention 2 or 3 individual 
 contents and to assign to them either the same or slightly dif- 
 ferent degrees, in other cases he would give merely their 
 average degree, while in a few cases he remained doubtful. 
 The last mentioned were omitted from the Table, likewise 
 those experiments in which the attention to the marking varied 
 between more than two neighboring degrees. In the case of 
 observers O and P several practice series were included in order 
 to get a suflScient number of results for tabulation. The Table 
 shows as definitely as can be expected under our experimental 
 conditions that the frequency-modes of distribution for the 
 simultaneous processes at first gradually approach, starting 
 with I and 7, then 2 and 6, then 3 and 5, then they fuse in 
 the 4th and 5th horizontal division, and finally they change 
 places with each other, occurring now at 6 and 3, at 7 and i, 
 at 8 and 1, and lastly at 9 and i. At the 4th and 5th degree 
 of marking, there is no mode distinguishable for the other 
 mental processes, because here we have a more or less uniformly 
 distributed attention without much of a focus or a background. 
 At the 6th level, however, the dual division becomes again 
 apparent and remains so for the rest of the Table. This 
 
 1 Pillsbury: Attention, 1908, 9. 
 
526 
 
 GBISStKR 
 
 Tabi,e VIII 
 
 
 
 
 
 
 Degrees 
 
 of attention 
 
 
 
 
 
 No. 
 
 Obs. 
 
 T. 
 
 I 
 
 2 
 
 3 4 
 
 5 6 
 
 7 
 
 8 
 
 9 
 
 Total 
 
 
 O 
 
 24 
 
 
 
 
 2 
 
 10 
 
 12 
 
 6 
 
 30 
 
 I 
 
 P 
 
 13 
 
 
 2 
 
 I 
 
 I 5 
 
 6 
 
 2 
 
 4 
 
 21 
 
 
 V 
 
 33 
 
 
 
 
 
 4 
 
 I 
 
 5 
 
 10 
 
 
 3 0's 
 
 70 
 
 0+70 
 
 2 
 
 I 
 
 I 7 
 
 20 
 
 15 
 
 15 
 
 71 
 
 
 O 
 
 8 
 
 
 
 I 
 
 I s 
 
 I 
 
 
 
 8 
 
 2 
 
 P 
 
 22 
 
 
 
 2 
 
 3 9 
 
 8 
 
 5 
 
 I 
 
 28 
 
 
 V 
 
 13 
 
 
 
 I 
 
 4 I 
 
 2 
 
 
 3 
 
 II 
 
 
 3 0's 
 
 43 
 
 
 0+43 
 
 4 
 
 8 15 
 
 II 
 
 5 
 
 4 
 
 47 
 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 P 
 
 13 
 
 
 
 2 
 
 7 6 
 
 4 
 
 2 
 
 I 
 
 22 
 
 
 V 
 
 14 
 
 
 I 
 
 2 4 
 
 I 
 
 
 
 7 
 
 15 
 
 
 3 O's 
 
 27 
 
 
 I 
 
 2+27 6 
 
 8 6 
 
 4 
 
 2 
 
 8 
 
 37 
 
 
 O 
 
 I 
 
 
 I 
 
 
 
 
 
 
 I 
 
 4 
 
 P 
 
 5 
 
 
 
 I I 
 
 3 2 
 
 I 
 
 
 
 8 
 
 
 V 
 
 14 
 
 I 
 
 2 
 
 7 3 
 
 I I 
 
 
 
 3 
 
 16 
 
 
 3 O's 
 
 20 
 
 I 
 
 3 
 
 5 4+20 
 
 4 3 
 
 I 
 
 
 3 
 
 25 
 
 
 O 
 
 3 
 
 I 
 
 3 
 
 
 I 
 
 
 
 
 5 
 
 5 
 
 P 
 
 5 
 
 
 
 4 
 
 I 
 
 3 
 
 1 
 
 
 9 
 
 
 V 
 
 16 
 
 5 
 
 4 
 
 3 3 
 
 I I 
 
 
 
 5 
 
 22 
 
 
 3 O's 
 
 J4 
 
 6 
 
 7 
 
 7 3 
 
 2+24 2 
 
 3 
 
 I 
 
 5 
 
 36 
 
 
 O 
 
 6 
 
 
 4 
 
 I 
 
 
 2 
 
 
 2 
 
 9 
 
 6 
 
 P 
 
 6 
 
 
 
 3 I 
 
 I 2 
 
 3 
 
 2 
 
 
 12 
 
 
 V 
 
 23 
 
 6 
 
 5 
 
 7 I 
 
 2 2 
 
 
 
 8 
 
 31 
 
 
 3 O's 
 
 35 
 
 6 
 
 9 
 
 II 2 
 
 3 4+35 
 
 5 
 
 2 
 
 10 
 
 52 
 
 
 O 
 
 6 
 
 2 
 
 2 
 
 2 1 
 
 
 2 
 
 3 
 
 I 
 
 13 
 
 7 
 
 P 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 26 
 
 19 
 
 2 
 
 4 3 
 
 2 I 
 
 
 
 12 
 
 42 
 
 
 3 O's 
 
 32 
 
 21 
 
 4 
 
 6 4 
 
 2 I 
 
 2 + 32 
 
 3 
 
 13 
 
 55 
 
 
 O 
 
 8 
 
 3 
 
 6 
 
 
 
 
 3 
 
 3 
 
 15 
 
 8 
 
 P 
 V 
 
 3 
 6 
 
 6 
 
 2 
 
 2 
 
 
 I 
 
 2 
 
 2 
 
 
 
 3 O's 
 
 17 
 
 9 
 
 8 
 
 2 
 
 
 I 
 
 S+17 
 
 5 
 
 30 
 
 
 O 
 
 3 
 
 
 3 
 
 I I 
 
 
 
 
 
 5 
 
 9 
 
 P 
 
 
 
 
 
 
 
 
 
 
 
 
 
 V 
 
 5 
 
 4 
 
 I 
 
 1 
 
 
 
 
 6 
 
 12 
 
 
 3 O's 
 
 8 
 
 4 
 
 4 
 
 I 2 
 
 
 
 
 6+8 
 
 17 
 
 mutual relation of the two clearness levels in an attentive con- 
 sciousness may perhaps be compared to the two levels of a 
 quicksilver column occupying about half the space of a U- 
 shaped glass tube. The one end of the column cannot rise to 
 its maximal height without depressing the other to a minimal 
 height, nor can the one end move toward a medium high point 
 
THE MEASURBMENT OF ATTENTION 527 
 
 without causing the other end to approach approximately the 
 same level, but of course from the opposite direction. This 
 analogy can apply, however, only to the two-level formation 
 of attentive consciousness, which since Leibniz has been known 
 under the terms apperception and perception. 
 
 This distinction has not been unanimously accepted by mod- 
 ern psychologists as a universally valid description for all pos- 
 sible states of attention. Some of the authors^ discover in their 
 own consciousness processes of more than two levels of clear- 
 ness, while Wirth even seems to hold that at any moment of 
 attention all possible degrees of apperception may be represented 
 by simultaneous processes and that only "under certain con- 
 ditions, which favor a kind of dual division, it is possible that 
 a region, to which a fairly uniform attention is given, may be 
 opposed to a 'background' or 'periphery' of consciousness, 
 to which attention is as uniformly denied' ' ." To this statement 
 we may reply that according to our results the occurrence of 
 the two-level formation in the attentive consciousness does not 
 depend upon "certain conditions", but upon the individual 
 observer. We believe we have sufficient evidence to justify 
 the assumption that there are two more or less distinct types 
 of observers, those for whom the two-level division is the most 
 natural and most common, and those who experience as a rule 
 several levels of clearness. 
 
 Of our 5 observers B, C, O, P, and V, who worked under 
 identical experimental conditions, the last three reported with- 
 out exception the dual division of clear focus and vague back- 
 ground. There is, of course, the possibility that during the 
 experiment, which lasted at least 12 seconds, processes of an 
 intermediate degree occurred but were too fleeting to be recalled 
 later. The first two observers, B and C, on the other hand, 
 frequently mentioned such processes on intermediate levels of 
 clearness. Here again it might be objected that our experi- 
 ments did not deal with a simultaneous consciousness, but with 
 successive states of attention, which might have exhibited a 
 variable maximum of clearness and thus, if viewed as a whole, 
 would present simultaneously what really was experienced in 
 succession. We can only offer in reply the introspective ma- 
 terial of our observers. We therefore quote a few passages 
 from their introspections: 
 
 Observer B 
 
 (i) Clearness of digits varied from 3-7, some attention was given to 
 the whole row, a few individual digits were as high as 3 in clearness. 
 Circles pretty dark, never higher than 5, and usually 8, while the 
 
 > Cf. Titchener's discussion of this topic, Lectures, 220-228. 
 "Wirth : Phil. Stud. XX, 493; cf. Amer. Jour, of Psychol. XX, I. 
 
528 GBISSLBR 
 
 marking was almost automatic, about 7-8. Other processes, e. g., 
 the noise of apparatus darkest, 8-9. 
 
 (2) Clearness of colors as whole rows 4-5, marking visually 5, as 
 motor adjustments perhaps 6. In the background mostly verbal com- 
 ments, 6-7. Sometimes flashes of colors, especially red and yellow, 
 rose occasionally as high as 3. 
 
 (3) Focal clearness was distributed among the circles which got 3. 
 Letters were quite varying, those next to the circles between 3 and 6, 
 others 8. The general conscious background was very dark and dead, 
 about 8 or 9, mostly verbal images, saying "I must avoid the letters", 
 or noise of apparatus, and the general experimental setting. 
 
 (4) Focus on average 3, mostly visual and kinsesthetic. The back- 
 ground as a rule 8, made up as usual. A few distractions, chiefly verbal 
 ideas, rose as high as 5 at times. 
 
 Observer C 
 
 (i) Marking of circles 90%. Individual colors not noticed, per- 
 haps s%. But the rows as such were confusing, the difficulty of finding 
 circles about 50%, being mostly strain in hand and eye accompanied 
 by verbal comment. 
 
 (2) Marking the circles about 80%, the digits not more than 10%, 
 they were scarcely seen. But the noise of the apparatus sometimes 
 rose as high as 40% and was markedly unpleasant. 
 
 (3) Marking 70%, colors about 30%, noise of apparatus 50%, rose 
 so high especially when it suggested ideas of railroad, which were 
 mostly visual and about the same in clearness. 
 
 It must be added that even with these two observers the dual 
 division was experienced at times; but it is significant that B, 
 previous to his participation in our experiments, had firmly 
 believed in the universality of the two-level formation. Among 
 the undergraduate students who took part in the prelimi- 
 nary experiments on clearness, we also found some insisting 
 that in their experiences during these experiments as well as 
 in their daily life the dual division was only rarely realized. 
 We must, of course, admit that in daily life we do not usually 
 analyze our consciousness, nor do we then sharply distinguish 
 between psychological clearness or prominence in conscious- 
 ness, logical clearness of the meanings and relations, visual 
 clearness or distinctness, and differences in intensity of simul- 
 taneous processes. Nevertheless our observers were aware of 
 these distinctions at least during the hours of experimentation. 
 Adding to this the fact of the difference of opinion among 
 expert psychologists upon thiis very point, we are led to make 
 the assumption of two types of attentive consciousnesses, the 
 dual division type and the multi-level type.* Whether these 
 two types are mainly a matter of individual difference of con- 
 stitution or of training and habit, or whether they depend upon 
 certain psychophysical conditions within the same individual, 
 must be left for experiments to decide. The distinction is not 
 
 1 Such a possibility is, if not admitted, at least implied in Titch- 
 ener's Lectures, 228. 
 
THE MBASURBMBNT OF ATTENTION 529 
 
 intended to be a substitute for Titchener's hypothesis that in 
 the two-level formation there may occur slight diflferences of 
 clearness both in the focus and in the periphery, but is offered 
 rather as a supplement to serve its purpose until a better ex- 
 planation is proposed to account for the differences of introspec- 
 tion regarding the number of simultaneous clearness levels in 
 the attentive consciousness. Our results furnish, as a matter 
 of fact, many instances for the coexistence of small clearness 
 differences on both the upper and the lower level of attention. 
 
 In conclusion we may briefly restate the general results of our 
 investigation : 
 
 ( 1 ) A very close parallelism was found to exist between 
 introspectively distinguishable variations of attention and cor- 
 responding differences in the precision of work performed at 
 these levels, under the condition that the estimation of degrees 
 of attention was made in terms of clearness and that the work 
 itself was not influenced by anything else but change in atten- 
 tion. 
 
 (2) Under the same conditions the introspective estimation 
 of the quality of the work was not as reliable as the evaluation 
 of the degrees of attention. 
 
 (3) It seems possible that by continued practice a differen- 
 tial clearness limen may be established which would be of great 
 assistance in the measurement of attention. 
 
 (4) The degree of concentration does not depend so much 
 upon the nature and number of distractions, as upon the nature 
 and complexity of two simultaneous tasks and the preliminary 
 instructions regulating Einstellung and direction of attention. 
 
 (5) There seem to be two types of the attentive conscious- 
 ness, the dual division and the multi-level formation. 
 
 (6) In the dual division type of attention a reciprocal rela- 
 tion exists between the two levels ; that is, the higher the 
 apperceptive level rises, the lower the perceptive level falls, 
 and conversely. 
 
_^ Cornell University Library 
 
 BF321 .G31 1909 
 
 Measurement of attention ... 
 
 oljn 
 
 3 1924 029 037 468