023-039Outline as a condition of attention

8/7/2019 023-039Outline as a condition of attention

1/17

OUT L INE AS A C ONDIT ION OF AT T E NT IONBY HERBERT WO0DR0W

University of MinnesotaObservations have recently been made in the psycho-

logical laboratory of Cornell University, concerning theeffect of form, or ou tline, upon clearness. I t was found byMeads1 that a definitely outlined light, in the form of across, had a greater power to attract attention than a light,of the same objective intensity, which appeared merely as a"spot" that faded off gradually into its black background.It was found, further, that the light in the form of a crosscould be considerably lessened in intensity and still attractas high a degree of atten tion as the formless spot. As acriterion of the ability of the stimulus to attract attention,introspective judgments of clearness were used; so that thestatement, that the cross attracted more attention than thespot, is to be taken as meaning that the subjects judged itto be clearer. M eads's work, then , may be regarded asexperimentally establishing the fact that definiteness of out-line is a condition of attention, in the sense that a visualstimulus with definite outline is judged clearer than onewithout it .The present study, though dealing with the same problemas that worked on by Meads, was made by a purely objectivemethod of measuring degree of attention, one that does notrequire introspections of clearness on the part of the subject.I have recently described this method in detail, and given,as it seems to me, an elaborate demonstration of its validity.

2This method seems of sufficient importance to justify testingit in every possible way, and adapting it to as many types of

1 'Form vs . Intensity as a Determinant of Attention, Minor Studies from thePsychological laboratory of Cornell University, ' communicated by E. B. Titchenerand H. P . Weld, Amer. J. of PsychoL, 1915, 150-151.

1 'The Measurement of Attention,' PsychoL Monog., 1914, Vol. XVII. , 5, pp. 158.2 3


2/17

2 4 HERBERT WOODROWproblem s as possible. T he problem of the effect of outlineis not only an interesting one in itself, but seemed betterthan any other for determining the degree of correspondencebetween clearness as judged introspectively and degree ofattention as measured objectively. On the whole, theresults obtained by Meads, though reported as a minor con-tribution, probably form the best introspective basis at handfor testing the validity of any objective method of measuringdegree of attention; and it was on this account that I under-took the present investigation. M y object was no t somuch to study the effect of outline upon degree of attentionas to ascertain whether the objective method, to which Ihave referred, can be given a further, introspective, justi-fication, by determining whether it leads to the same con-clusions as are reached by means of introspection on the partof trained observers. Pillsbury writes th at " . . . the onlytest of a method is its accomplishment. . . . Any methodthat gives results must be kept." 1 It will be seen from thedata below, that both the objective measurements herepresented and the systematic introspections reported byMeads result in the same conclusion, namely, that definite-ness of outline is a condition of attention.

The objective method used is one employing reactiontimes. I t consists in first ob tain ing a series of reaction timeswith a regularly repeated preparatory interval of 2 sees.,and then in obtaining a series of reaction times with a set ofirregular preparatory intervals of widely differing lengths.The reciprocal of the absolute prolongation produced by theirregular preparatory intervals is taken as the measure of thedegree of attention given to the stimulus, or rather to thechange of stimulus, to which the subject reacts. T he theo ryof the method is similar to that of the well-known distraetionmethods that have been proposed for the measurement ofatten tion. Th e unfavorable prepa ratory intervals are re-garded somewhat as a distracting influence, though in viewof the fact that they merely constitute an obstacle to a high1 "The Function and Test of Definition and Method in Psychology,' Science, VoL

XLI. , 1915 .1054 , 375 .


3/17

STUDIES ON ATTENTION 25

degree of attention, without causing a division or tearingapart of attention, it seems preferable to speak of them as adetracting influence. The absolute prolongation in reactiontime which is produced by the use of the irregular p reparato ryintervals may be spoken of as the absolute detraction effect.I have previously shown that this absolute detraction effectvaries inversely as the degree of attention given to the changein stimulus to which the reaction occurs.1 The reciprocal ofthe absolute detraction effect may therefore be taken as ameasure of the degree of attention.Now a reaction is always a reaction to a change, and theattention measured by the method just described is, therefore,always the attention to some change. Hence, to study out-line as a condition of attention, it is necessary to determinewhether a variation in the definiteness of outline of a stimuluscauses a variation in the degree of attention to some changein that stimulus. If the degree of attention becomes greateras the stimulus becomes more definite in outline, we mayexpect that, at the same time, the degree of attention to agiven change in the stimulus will likewise become greater.2The change used, in most of the work, and in all that reportedin Tables I. to IV., was a slight decrease in intensity.My procedure for studying the effect of definiteness ofoutline as a condition of attention, then, was as follows:Two visual stimuli were used, both of the same objectiveintensity. One was a definitely outlined square of light,the other, a formless spo t. W ith each of these stimuli,reaction times to a slight decrease in intensity were obtained,both with regularly repeated 2 sec. preparatory intervals andwith a set of preparatory intervals varying from 4 to 20 sees,and mixed in irregular order. The reciprocal of the absoluteprolongation in reaction time produced by the irregularpreparatory intervals is taken as a measure of the degree ofattention to the decrease in intensity. If the degree ofattention thus found is greater with the definitely outlined

lOp.dt.,Pu.p.lVL* See Wirth, Philot. Stud., Vol. XX., 1902, 625-635. Also Woodrowr, op. cit.,

118.


4/17

2 6 HERBERT WOODROWsquare than with the spot, it means that definiteness of out-line is a condition of attention.The intensity of the stimulus used was quite weak, sinceit was found that the difference in clearness between the spotand square was then much more pronounced than when theintensity was great. Th e change in intensity, it is im po rtan tto note, was produced at the source of illumination, a stere-opticon lamp, and therefore affected the entire square andthe entire spot. As some of the subjects could notice amuch smaller change in intensity than others, different sizeddecreases were used with different subjects. In each case,preliminary trials were made to determine the size of de-crease just great enough to be always promptly noticed, evenin the case of the spot, with irregular preparatory intervals.When the size of decrease suitable for a given subject wasonce determ ined, it was no t thereafter changed . I t wasfound th a t as the size of the change in intensity was increased,the effect of outline became less im po rtant. However, evenwith the subjects who had the finest discrimination, subjectsLI and Ww, the largest size of change used, that with subjectBn, was still affected by outline.Before deciding to use a decrease in intensity as the changein stimulus to which the subject should react, I made pre-liminary trials, with several of the subjects, in which I used aslight increase, and also the complete cessation and thesimple occurrence of the light. W ith some subjects, pro -vided a light of weak intensity was used, the degree of atten-tion to an y of the four types of change was marked ly affectedby definiteness of ou tline. However, in the case of the mereoccurrence or the complete cessation of the stimulus, or evenwith a very large increase or decrease in intensity, unless thestimulus was very weak, the effect of outline upon the clear-ness of the change was com paratively slight. This w asdetermined not only by introspection but by a number ofmeasurements of the degree of attentio n. W hen the mereoccurrence of the light was used as the reaction stimulus,even with a weak light, one of the two subjects tried showedabsolutely no effect as the result of outline, while the other


5/17

STUDIES ON ATTENTION 27showed a marked effect. The subject who showed no effectgave the introspection that he reacted before he becamedefinitely aware of the outline, he merely knew that "some-thing had happened ou t there, and apprehended it as a definitesquare or as a spot only after reac ting." Th is introspectionwould suggest that an awareness of the form of the outlineis first necessary if it is to exert a favorable influence on thedegree of atte ntio n. Fu rther investigation, however, withvery weak intensities of stimulus, and with more subjects,would be necessary to establish this point.The subjects were carefully instructed to make everyreaction as quick as possible. Only with such instructionsis there any guarantee of the soundness of the method used,for then, only, is the time required for the reaction an ade-quat e m easure of the subject's efficiency. As the stimuliwere always in view, no special fixation point was provided.The subjects merely fixated as nearly as possible the center ofthe stimulus, whether square or spot.In all, seven subjects were used. W ith two of these, aseries of 10 measurements were made, for both square andspot. W ith the others, as man y measurements were madeas their time perm itted. A t each sitting, 50 reactions weretaken with both the square and the spot, 25 with a regularlyrepeated interval of 2 sees., and 25 with irregular preparatoryinterva ls. T he order in which the square and spot were usedwas changed each day. T he set of 25 irregular pre pa rato ryintervals was the same as I have used in previous work,namely, 4, 16, 20, 4, 8, 20, 12, 16, 8, 4, 12, 4, 20, 8, 12, 4, 16,20, 8, 16, 12, 12, 8, 16 and 20 sees. The order of the firstfew intervals was changed occasionally with those subjectswho were used oftener than two or three times.

Details of Apparatus.The lights used as stimuli were projected from a stereop-ticon onto a ground glass partition 180 cm. square, built into the wall separating thesubject's room from the experimenter's room, both of which rooms were dark-rooms.The only illumination in the experimenter's room was a small shaded lamp, by whichto read th e chronoscope. The light from th is lamp was prevented from falling ontothe projection glass by a curtain, which entirely screened off the experimenter's endof the room. Between this curtain and the projection glass, was placed the stereop-ticon, the objective of which came about 75 cm. behind the glass. Th e stereopticonwas illuminated by a frosted-tip, 40-wa tt, M azda, tungsten lam p, in series with another


6/17

2 8 HERBERT WOODROW40-watt Mazda, on a 115-volt D .C . circuit. Th e intensity of this light was veryconsiderably reduced by means of a yellow and a blue-green sheet of gelatine, placedin the path of the projected ray s. Th e voltage for the lamp circuit was supplied fromthe university power-house. A Weston voltmeter was kept constantly connected w iththe circuit terminals, and placed right behind the chronoscope, where the experi-men ter could watch it. M ost of the time the voltage was fairly steady. On certaindays it was not, however, and work was given up on those days. Through out t hegreat majority of sittings, there was not a variation of more than 2 to 4 volts, and nosudden jum ps.

In order to obtain a spot of light without definite outline, the stereopticon objec-tive was thrown com pletely out of focus, and no slide put in the slide-holder. Thesubject's room was dimly illuminated by a 10-watt frosted Mazda placed about 5: meters behind, and 2 meters above, the subject's head. Th is dim illumination wasused to make it easier to secure perfect retinal adaptation, as it is well known that avery long and variable time is required for complete adapta tion to darkness. W ith' this illumination of the subject's room, the spot of light appeared somewhat yellowish, green, bu t very poorly satu rate d. I t differed bu t little in color tone from the rest of; the ground-glass plate which formed its background. I t is difficult to give the dimen-sions of the spot, as it shaded off so gradually that it was impossible to tell where it, ended. Its just visible edge was abou t 18 cm. in diameter. I t was jus t noticeably5 less bright at 4 cm. from the center tha n at the center.The square of light was obtained merely by lowering a dark gray paper screen onthe opposite side of the projection plate from the subject. In th e center of this screen; was cut a square 7 x 7 cm. Th e change from spot to definitely outlined square wasH made, then, simply by cutting off th e edges of the spot. The reason for using asquare, instead of following Meads in the use of a cross, is that a square was foundto present a surface of more uniform brightness than a cross of the same area. Along, the border of the square one could discern a narrow contrast rim, 2 or 3 mm . wide, of' heightened intensity. Th e brightness of the screen itself was so chosen th at , uponi being lowered into position, it did not change the apparent intensity of the background.r Th e decrease in brightness to which the subjects reacted was produced by closing1 a circuit in parallel with the stereopticon lam p. Th is circuit contained a series offour Mazda lamps, which were not the same for all subjects. By varying the resistanceof these lamps, the amount of decrease in the intensity of the stimulus could be varied.As already stated, a decrease in intensity which seemed just sufficient to be alwayspromptly noticed by the subject was chosen by preliminary trials with each subject.The circuit in parallel with the stereopticon lamp was closed simultaneously withthe separate chronoscope circuit, by m eans of a double key. The chronoscope, aHipp 's, was arranged in series with a storage b atte ry of abo ut 12 volts, and controlledbefore and after each sitting by the Wundt fall-hammer, set for a time of 201 a. The. warning signal used was the click of an electric sound-ham mer. Th e dura tion of thepreparatory intervals was judged by the experimenter with the aid of a head telephonereceiver, connected with an electric metronome, beating seconds, in an adjoining room.The results for each subject are given in Tables I. and II.,and a summary of the results for all subjects in Table III.The reacti on times obtained w ith the definitely ou tlined squareare given in the columns headed 'Square,' and those with the


7/17

STUDIES ON ATTENTION 29indefinitely outlined spot in the columns headed 'Spot.'The columns headed '2 sees.' and those headed 'Irreg.' givethe average reaction times with regularly repeated 2 sec.preparatory intervals and with the set of irregular prepara-tory inte rva ls. T he difference in the average reaction timewith the 2 sec. intervals and with the irregular intervals isgiven in the columns headed i/A and i/A'. Since thisdifference varies inversely with the degree of attention, Astands for the degree of attention given the decrease inintensity in the case of the square, and A' for the same inthe case of the spot. The figures in italics are the m eanvariations of the averages above them.A study of Tables I. to III. shows that, in every case,there was a marked prolongation in reaction time producedby the irregular preparatory intervals, but that this pro-longation was always greater with the spot than with thesquare. Now the prolongation produced by the irregularpreparatory intervals varies inversely as the degree of atten-tion. Therefore, since a t every sitting this prolongation isgreater with the spot than with the square, it follows thatthe degree of attention to the change is less in the case ofthe spo t than in the case of the square. T he differencebetween the two cases in the amount of prolongation pro-duced by the irregular intervals indicates the difference inth e degree of attention. This difference is given in thecolumns headed ijA' i/A. In Table III., I have addedanother column, headed (A A')jA, which, upon theassumption that the measurements of degree of attentionmade by the present method may be regarded as absolutemeasurements, gives the percentage of decrease in degree ofattention produced by using the spot in place of the square.The correctness of the general conclusion that the degreeof attention to the change in intensity is less when it occursin the spot than when in the square, is strongly confirmed bythe sub jects' introspections. These introspections were verycarefully noted throughout the work, as it was mainly thepossibility of obtaining a good introspective confirmation ofthe method of measuring attention here used that led to the


8/17

3 HERBERT WOODROJF

TABLE ITHE EFFECT OF DEFINITENESS OF OUTLINE UPON DEGREE OF ATTENTION TO A SLIGHT

CH ANCE IN INT E NS IT Y

Subj,

Wwu

u

ft

fi

A vAv. M. V. .LI

tf


9/17

STUDIES ON ATTENTIONTABLE II

A CONTINUATION OF TABLE I

Subj.

Fstt

AvAv. M.V .. .St(I

tt

AvAv. M.V. . . .Bstt

tt

tt

tt

It

tt

AvA v . M . V . . . .Dgit

AvA v . M . V . . . .Bntt

AvA v.M .V . . . .

> Sees.341228219293162802029530297342882029O5*263202592223221247222352742492725125295352822028928332555 < 53 I 134752

SquareIrreg.3933238552389324134738842385553794.1391*J38554413273705*373353692538s573565037834426504074175'460504784906247655

M8910396

133938891101

122

154138126134ill107127

131

125

128

128103157129

Spot2 Sees.

3"262922030223296353035*305503173830538282242645*25527268252552(52882*2S6202672*3225"2992831052374554103967039361

I r r e g .

486624364246152479(5/441984275*4656845371474544854943834449So4355


10/17

3 2 HERBERT WOODROWTABLE III

A SUMMARY OFTA BLES I AND II

Subj.

W w . . .LIFs . . . .St . . .Bs . . . .D g . . . .B n . . . .

Square*Secs.

260461929316290312512328928347

Irreg.337353703895239141378344175*476S3

llA779496

IOJ127128129

Spot3 Sees.2712 *288243022J3SA283105239361

Iirep.379

ss461524S3&s s5460995

i M '108143159148189196216

z/A'i/A

31496347626887

AA'A

.283439.32333S4

p resen t inves t iga t ion . All the s u b je c t s , w i th o u t e x c e p t io n ,gave the i n t ro s p e c t io n th a t the c h a n g e was clearer and eas ie rto not ice when i t o c c u r r e d in the s q u a r e . The sub jec t s werea l l jus t as s u r e t h a t the c h a n g e in the s q u a r e was c lea re r thanthe change in the s p o t , as t h e y w e r e t h a t t he square i t6elfwas c learer than the s p o t .

The following are a few t y p i c a l i n t ro s p e c t io n s :L I : " T h e s q u a re s t a y s in v ie w b e t t e r , and the effort of

c o n c e n t r a t i o n is less . Af ter wo rki ng w ith the s p o t , I a l w a y ssee the square come wi th a sigh of relief. A lw a y s t i r e d w h e nthrough with the s p o t , bu t not af ter the s q u a r e . W i n k i n gd i d n ' t b o t h e r in the case of the s q u a r e , but was n e c e s s a ryw i th the s p o t to re l ieve s t ra in of c o n c e n t r a t i o n . T h e c h a n g ein the s q u a re is c e r t a in ly c l e a r e r t h a n th a t in the s p o t . "

B s : " T h e s q u a re is m u c h c l e a r e r e a s i e r to see. Thechange is easier to no t ice in the case of the s q u a r e n o d o u b ta b o u t it. The c h a n g e is so m u c h m o re d i s t i n c t in the caseof the s q u a r e t h a t i t gives me m o r e of a c h a n c e t o t h i n k - a b o u tr e a c t in g q u ic k ly . "

B n : " C h a n g e t w i ce as e a s y t o n o t i c e w i th s q u a re asw i th s p o t . "St: "There is no doubt about i t" (that the change in thesquare is clearer than that in the spot).Dg: "Square is of course more distinct and somewhat


11/17

STUDIES ON ATTENTION 3 3brighter, but it is also clearer, that is, makes more of animpression on you . T he change is much clearer in thesquare."In addition to such introspections as those noted above,systematic observations were made by myself and myassistant1 on the effect of varying the absolute intensity ofthe light and also of varying the size of change in intensityof stimu lus. We both observed tha t, with increase in theintensity of the stimulus, there occurred a marked decreasein the effect of outline upon the clearness of the stimulus,and, similarly, with increase in the size of the change inintensity, there occurred a marked decrease in the effect ofoutline upon the clearness of the change. T he correctnessof these observations was checked by objective measurementsof degree of attention made with a high intensity of stimulusand a very large decrease in inten sity. These measurem entsshowed a high degree of attention to the change in intensity,in both the square and the spot, but no reliable differencebetween the two.In interpreting these observations on the effect producedby variation in intensity and in size of change upon theimportance of definiteness of outline, it should be rem emberedthat just as the degree of attention to a simple stimulus isconditioned objectively by the intensity of the stimulus, so isthe degree of attention to a change in intensity of stimulusconditioned by the size of the change.2 Now, since theeffect of outline becomes less marked as the size of changeor the absolute intensity is increased, it is evident that theeffect of outline as a condition of attention becomes less asthe degree of attention affected becomes greater. Forexample, if the square and the spot both have a very greatbrightness, the degree of attention which each attracts ishigh, and there is then little, if any, difference in clearnessbetween them . If, on the other hand, they are both ve rydim, the square is much the clearer. Similarly, as regardsthe effect of outline on the degree of attention to a change in

1 Assistance in this study, as well as in several others, was made possible by aidreceived from the research fund of the University of Minnesota.

1 Woodrow, op. cit., 114.


12/17

3 4 HERBERT WOODROWintensity, if the change is large, it attracts an equally highdegree of attention in both the square and the spot, but ifit is small, it attracts much more attention in the square.The fact that the effect of outline upon degree of attentionbecomes less important with increase in the degree of atten-tion otherwise conditioned, is merely an illustration of thegeneral law, that "the absolute effect of a given detractor ofattention varies inversely as the degree of attention uponwhich the detractor acts." 1 The detractor of attention, inthe present instance, is a given degree of indefiniteness ofoutline. Another well-known illustration of this law lies inthe fact that an "attention wave" is easily discernible onlyin the case of a very weak stimulus, such as produced by therotating Masson's disk, in which case the degree of clearnessis also very low because of the insufficiency in the objectivecondition of inte ns ity. Here we are evidently dealing w ithsome detracting influence, of a physiological nature, whichhas a marked effect when the degree of attention affected islow, bu t little or none when it is high. I have elsewhere2described in detail a number of other illustrations of this law,using irregular preparatory intervals as the detracting influ-ence.One might ask why the simple reaction times with theregularly repeated 2 sec. intervals could not be used aloneas a measure of degree of atte ntio n. Perhap s they could,in the present instance, but they would not serve as a verysensitive measure and one could not be certain that thedifference in simple reaction time between the reactions tothe change in the square and those to the change in the spotwas due to the difference in degree of at tent ion. The reactionprocess is complex, and variation in simple reaction time maybe the result of variation in other factors than attention.Difference in the prolongation produced by the irregularpreparatory intervals, however, can be due only to differencein the degree of attention in the two cases, since this pro-longation is due solely to the effect of the set of irregular

1 Op. cit., 99.lOp.cit., 157.


13/17

STUDIES ON ATTENTION 3 5intervals as a detractor of attention.1 Moreover, by usingthe reciprocal of the absolute prolongation in reaction timeproduced by the irregular intervals, we obtain a measure-ment which is comparable with other measurements of thedegree of attention, made at other times and under othercircumstances.To make an accurate comparison of the absolute effectof outline upon two different individuals, it would be neces-sary first to equalize their degree of attention in the case ofthe square. Th is could be done directly, by vary ing the sizeof the change to which they were to react, until the prolonga-tion produced by unfavorable intervals was the same for both.This would consume a great deal of time, but by using severaldifferent sized changes, one could easily calculate the sizesof change needed to make the degree of attention of the twoindividuals the sam e. After having equalized the degree ofattention for the two individuals to a change occurring inthe square, we could then measure the degree of their atten-tion to the same change occurring in the spot, and thusdetermine whether the effect of outline as a condition ofatten tion was the same for both . N o accura te measurem entof individual differences has been attempted by the methodjus t described, bu t there would be no great difficulty in makingsuch. The data , as obtained, indicate marked individualdifferences but it is certain t h a t these would be greatlylessened if the subjects were made equal in degree of atten-tion, since, in general, there is evident a considerable degreeof correlation b etween the reciprocal of the degree of atte ntionto the change in the square (i/A) and the unfavorable effectof indefiniteness of outline upon the degree of attention(i/J' - i\A).T he fact, th a t a fixed change in the inten sity of a stimulusattracts more attention when the stimulus is definitely out-lined than when it is not, can be due only to the favorableeffect of definiteness of outline, as such, upon the degree ofatten tion . I t is tru e th a t the square had a slightly greateraverage apparent intensity than the spot, presumably as the

1 Op. cit., 49-63 and 114.


14/17

3 6 HERBERT WOODROWresult of contrast. Th is fact, however, would no t accountfor the greater clearness of the change in the case of thesquare, for the effect of contrast is to decrease the notice-ability of a slight change in intensity . Anyone may readilyconvince himself that this is the case, by observing a verysmall change in the intensity of a light when it is matchedwith its background in brightness, and then observing thissame change when the background is either markedly lighteror darker. In th e latte r case, this change is less clear.1After all, so long as enough contrast to give a definite outlineis present, variation in degree of contrast has very littleeffect upon th e clearness of a given relative change in in ten sity.

In order to determine beyond question, however, whetherthe greater degree of attention to the square was due to anyincrease in apparent intensity apart from its greater clear-ness, an additional experiment was m ade. In this experi-ment, by using a background considerably brighter than thestimulus, the square was rendered definitely less in apparentintensity than the spot. Nevertheless, the square wasdefinite in outline and was judged unquestionably clearer thanthe spot. This may seem to run counter to the state m entth a t intensity is an im po rtan t condition of atte ntio n. As amatter of fact, it merely illuminates the meaning of thatsta tem en t. In the case of light, no t intensity , but differencein intensity, is the real, intensive, condition of attention.

2An experiment was conducted, then, in just the same wayas the work already described, except that the definitelyoutlined square was surrounded by a brighter background,and therefore rendered slightly, but unquestionably, less inappare nt intensity than the spot. The definiteness of ou t-line of the square was judged by the two subjects used in

I This seems to be due to the fact that the effect of contrast, relative to the differ-ence in intensity of the contrasting surfaces, is greatest when the difference is verysmall. See Lehm ann, 'Ueber die Anw endung der mittleren Abstufungen auf denLichtsinn,' PhUos. Stud., 1886, Vol. III. , 516-528.

I 1 have found by systematic introspection on the part of myself and subject LIthat the increase in clearness of the stimulus, with increase in the difference in intensitybetween the stimulus and its background, is marked when the difference is slight butbecomes much less noticeable after a considerable difference has once been reached-This holds, no matter whether the background is lighter or darker than the stimulus.


15/17

STUDIES ON ATTENTION 37this experiment to be about equal to that of the square withthe dark background used in the work already presented.The same source of illumination and the same physicaldecrease in intensity were used in this experiment as in theprevious work with these subjects, reported in Table I.The data are given in Table IV., where the headings of thecolumns have the same significance as in the preceding tables.

TABLE IVT H E E FFE C T , U PON T HE D E G R E E OF A T T E NT ION T O A SU GH T D E C R E A SE IN INT E NSIT Y ,

OF DEF INI TEN ESS OF OUTLIN E PRODUCED BY SURROUNDING TH E STIMULUSWITH A BRIGHTER BACKGROUND

Sttbj .Square

Sees.2093 83023229O292 9S2629731

297l83012126318265233061738619

I m g .3763 63823 6367327*3763 6

4073 7381463693 837126378253813 4

i/ A7 780778479no801061067295

Sp o tx S e e s .

3OO203123 3308402942230429

304252972228S23283163O62229522

Irreg.

4166 34O43 3398253943 44033 9

4843 64264~42 0223 97254154'4383 4

ilA>

116929010099180129135114109133

ilA'i/A

Ww.

AvAv. M. V.LI.

AvAv. M . V.

39121316

704929

83538

From a comparison of the values of ijA in Tables I.and IV., it is evident that the degree of attention to thechange in intensity in the case of the square was affectedvery little, if at all, by using a bright background in place ofa dark one. For subject Ww , with the dark background(Table I.), the average value of ijA is 77, and with the lightbackground (Table IV.), 79; while for subject LI, the corre-sponding values are 94 and 95 . In th e case of the spot, someimprovement in the degree of attention is to be noted in


16/17

3 8 HERBERT WOODROWTable IV., as the value \\A', for subject Ww, changes from108 in Table I. to 99 in Table IV., while for subject LI, itsvalue changes from 143 to 133. Since the reactions with thespot given in Tables I. and IV. were obtained under identicalconditions, the difference is due either to 'accidental varia-t ion / or, since the results of Table IV. were obtained afterthose of Table I., to practice. While I have never been ableto obtain any convincing evidence that the detracting effectof unfavorable preparatory intervals wears off with practice,there is no reason why practice might not lessen the effective-ness of such a detractor of attention as indefiniteness ofoutline.

From the data of Tables I. and IV., the conclusion follows,tha t the greater clearness of the change in the square is notdue to the effect of contrast on the apparent intensity of thesquare, but to the greater clearness of the square, which inturn is due to the definite ou tline. Definiteness of outlinemean6 that there is a definite setting off of the stimulus fromits background, and it is this fact alone that gives the squarethe greater clearness. Definiteness of outline, though itmay involve brightness contrast, is thus a condition ofattention apart from the effect of contrast on apparentintensity.CONCLUSIONS1. The effect of outline up on the degree of attention givento a visual stimulus may be measured by measuring its effectupon the degree of attention to a small change in that st imu-lus, preferably a small decrease in intensity. The degree ofattention to any change which may be used as a reactionstimulus is measured by the reciprocal of the absolute pro-

longation in reaction time produced by using a set of un-favorable preparatory intervals in place of 2 sec. preparatoryintervals.2. Measurements made by this method show that definite-ness of outline is a condition of attention. This statementmeans that the reaction time to a slight decrease in theintensity of a definitely outlined square is less affected by


17/17

STUDIES ON ATTENTION 3 9unfavorable preparatory intervals than is the reaction timeto the same decrease in intensity in an indefinitely outlinedspot.3. The results obtained by the purely objective methoddescribed above agree with the results of systematic intro-spection on the part of trained observers.4. The effect of definiteness of outline as a condition ofattention is not due to the effect of contrast on apparentbrightness. The essential condition upon which the favor-ableness of outline depends is merely that the stimulus shalldiffer sharp ly from its backg round . I t may be either lessintense or more intense.5. The effect of definiteness of outline upon clearnessdecreases with increase in the absolute difference between theintensity of the stimulus and that of its background.6. The effect of definiteness of outline upon the clearnessof a change in intensity decreases with increase in the sizeof the change.Conclusions 5 and 6 are merely particular instances ofthe general law that the absolute detraction effect exerted bya given detractor of attention varies inversely as the degreeof atten tion detracted from. In the present instance, thedetractor is a given degree of indefiniteness of outline, theeffect of which becomes small when the degree of attentionotherwise conditioned is high. T he degree of atte ntio n tendsto be high, in the case of attention to a stimulus, when thedifference in intensity between stimulus and background islarge, and in the case of attention to a change in stimulus,when the size of the change is large.

Documents

023-039Outline as a condition of attention