advertising visual vs auditive

7/27/2019 Advertising visual vs auditive

1/22

MEMORY FOR ADVERTISING COPY PRESENTEDVISUALLY VS. ORALLYFRANK N. STANTONi

The Ohio State UniversityThe present experiment was designed to determine thecomparative effectiveness of fictitious advertising copypresented to the subjects by the printed page versus thatdelivered through a radio loudspeaker. The method inmany respects follows that used by DeWick.2The advertising copy was presented by loudspeakerwithout a program in the sense of music, dialogue, etc.,

while the printed material was placed before the sub-jects void of illustrations, display types and other atten-tion arres ting devices. The effort was to eliminate asfar as possible all variables but visual versus auditorypresentation.The advertising copy used for both auditory andvisual presentation consisted of two groups of eightadvertisem ents each, hereafter denoted, Copy X and CopyY. W ith one group of subjects for instance, Copy Xwas given visually and the Copy Y by loudspeaker. Withanother group X was auditory and Y visual. The com-modities in each series represented articles frequentlypurchased, universal in use, and low in unit cost. In-cluded in each list there was one product of especial in-terest to men and another especially attractive to women.

1 The auth or wishes to acknowledge th e kind cooperation ofDr. Harold E. Burtt of The Ohio State University who contributeduseful and constructive criticisms. Dr. F . H. Lumley also ofThe Ohio State University assisted by helpful advice. DeWick, Henry N., "The Recall Value of Visual and AuditoryAdvertising Material" (unpublished).

45


2/22

46 FRANK N . STANTONThe remaining six products were presumably purchasedequally by both sexes.The fictitious trade names selected for the 16 com-modities in the two lists were two syllable names chosenon the basis of a preliminary study with visual andaud itory presentation . Confusing final consonants suchas s and f were distributed equally in the two lists. Notwo names ended with the same syllable or syllables thatsounded identically. The same precaution was main-tamed for the first syllable of each word. The two listswere balanced for similar products wherever necessary,e. g. Gasoline vs. Oil, Tea vs. Coffee. The two lists oftrade names and their products are printed below.

X SEEIES Y SERIESLorenz Powder Barlow CigarettesHarvey Toothpaste Par ke r CoffeeBry an Gasoline Crawford BreadW arren Candy Stew art OilKenwood Soap Bobbins Shav ing CreamDouglas Cigars Andrews GingeraleCurtis Ink Browning HosieryRandolph Tea W agner Mouthwash

A further attempt was made in the preliminary seriesto insure equal trade-name-commodity-associations. Agroup of 40 subjects, similar to those used in the ex-periment proper, participated. The experimenter reada list of fictitious trade names to the group with theinstructions to write the name of any product they couldassociate with the trad e nam e. After collecting thoseresponses the experimenter repeated the fictitious tradenames and with each one gave a product which he hadarbitra rily selected to go w ith it. To all subjects th ispresentation was made orally. A portion of them re -ceived typed copies of the fictitious trade name com-modity lists which they read as the experimenter pre-sented the list verbally. Imm ediate recall and aided


3/22

MEMORY FOR ADVERTISING 4 7recall tes ts were given. After an analysis of both setsof responses, those trade names for which there existedany consistent product associations other than the ficti-tious ones of the experiment were eliminated. Likewisecases where the fictitious trade-name-commodity-associa-tions ran extremely high or low with the experimentalgroup were eliminated. The effort was to select materialfor which there was no naturally strong association be-tween product and brand and also in which the forma-tion of such associations was not extremely difficult.Each piece of advertising copy was of approximatelyconstant lengthbetween 70 and 75 wordsand in-cluded trade name product mentions thre e times. Thefirst mention occurred at the opening, another approxi-mately in the middle with the third at the close of theparagraph. In each mention the trad e name alwayscame first with the product immediately after. Carewas exercised to keep all copy as near equal in com-plexity and sales appeal as possible. No dramatic state-ments or slogans were employed. A sample piece of copyis presented below.

"F or better times, Andrews Gingerale. A pleasa nt dinner, anenjoyable evening and then the grand finale. Only the sp arkleand zest of this famous pale dry Andrews Gingerale will makeyour pa rty complete. The new larg e bottles hold enough forfive full g lasses , making anyw here from five to eight long, tal ldrink s. And the last glass out of the bottle is ju st as good asthe first. Andrews Gingerale holds its lif e "

There were 8 such pieces in the X and Y sets16in all.Subjects for this investigation were 160 students ofboth sexes enrolled in psychology courses at The OhioState University during the summer quarter of 1933.They were divided into four groups to permit the neces-sary reversals of copy and method of presentation.Table I shows how copy and method of presentationvaried with the 4 groups. Fo r instance, Group A had


4/22

48 FRANK N. STANTONCopy X auditory followed by Copy Y visual at the samesitting-. TABLE I

Groups

Copy Used and Methodof Presentation

Number of Subjects

AX

audYvis39

BY

audX1 vis

38

CXvisYi aud

42

DYvisXaud41

The visual presentation was effected by the use ofbooklets. These measured 6 inches in width and 9 inchesin height and used an antique India3 stock for the textwith a cover of dark blue pap er. The instruc tions andfictitious advertising copy were set in 12 point, bold,antique face,4 18 picas, leaded two poin ts.3 No emphasiswas given to trade names or products by the use of caps,italics or special typ e. Copy only occurred on every fourthpage making it necessary to use 40 pages in each booklet.The cover bore the word "Front" set in 48 point Goudybold with a double rule under the word on the X seriesand a single rule under "Front" on the Y series.The same copy as prepared 6 for the booklets was used

3 Ind ia Mayflower Text, 60 lb. basis, Central Ohio Pa pe r Co.,Columbus, Ohio.4 12 pt. bold antique, designed for use with DeVmne, light face,Mergenthaler Linotype Company, Brooklyn, N. Y.s Prin tin g specifications followed in general the findings ofPaterson, Donald G., and Tinker, Miles A., "Studies of Typo-graphical Fac tors Influencing Speed of Readin g": VII I. "SpaceBetween Lines or Leading." Journal of Applied Psychology 1932,16,388-397. An electrical transcription of the auditory presentation wasprep ared on a Fairch ild Professional Record. This was not usedin the experiment.


5/22

MEMORY FOE ADVERTISING 4 9for the auditory presentation. For this purpose asmall portable public address system and a student an-nouncer were used. The public address system had amaximum output of 7 watts . W ith it were used a double-button microphone and a dynamic loudspeaker. Theannouncer read the copy to the groups. Several pre -sentations were made with the announcer and experi-menter for practice in order to eliminate possible stressor emphasis on any point. The trade names were notspelled out. The announcer and amplifying un it wereoutside the classrooms in which the auditory presenta-tions were made.

The following directions were given by the experi-menter when the visual presentations came first (GroupCandD, Table I ) ."Shortly , I shall pass out some booklets to you. Please place

them so that the word 'Front' on the upper half of the coveris facing you.7 When I give the signal you will please breakthe seal and tu rn at once to the first prin ted page. After youhave finished reading it please turn to the next page which willbe blank and then look up to receive the st ar tin g sign al. Ready. . . . begin."Further instructions for the visual part of the experi-ment were presented on the same paper with the same

type as the advertising copy in order to facilitate adjust-ment of the subjects for the copy to come. The instruc-tions read by the subjects were as follows:"On the following pages you will read a series of advertisementsfor fam iliar products Will you please sit comfortably and readcarefully the copy. La ters you will be asked to answer some7Each copy of the booklet had a 1% " band of wh ite pa perwrapped around it upon which these instructions were printed:NOTICE: Do not open this booklet until you have been instructedto do so.s Realising th at some subjects would suspect subsequent tes tingfor the content of the ads we felt it would be wise to spread theirattention as much as possible over all the copy by the generalstatement noted.


6/22

5 0 FRANK N . STANTONquestions about the products and the trade namesthe featuresof the prod ucts and the inducements to buy. Time enough willbe allowed to read each advertisem ent throu gh once. W henyou have finished reading the ad please turn immediately to thenex t page in this booklet. The pag e will be blank . After youhave done so await the signal from the experimenter beforetur nin g to the next advertisement. The time between rea dingswill give you an opportunity to think over the ad just read.Throughout the experiment will you please remain quiet so asnot to disturb your neighbor."

Before giving the signal to begin the experimentergave an opportunity for questions but none were askedin the en tire experiment. Following th at he gave thesta rtin g signal. At th e end of each succeeding minutethe signal to turn was given until all eight advertise-ments were read. The booklets were collected and aftera pause came the following announcement through theloudspeaker which had been installed before the classconvened:

"In the following minutes you will hear a series of advertise-ments for fam iliar p roducts. W ill you please sit comfortablyand listen carefully to the copy. Later you will be asked toanswer some questions about the products and the trade namesthe features of the products and the inducements to buy. Betweeneach advertisement there will be a pause during which you willhave time to think over the ad ju st heard. Throughout theexperiment will you please remain quiet so that you will notdisturb your neighbor?"

After those instructions the experimenter, still pres-ent, asked for questions and when the class was readyhe so signalled the announcer who began to deliver theauditory series over the public address system. The timeconsumed in delivering each piece of copy ran approxi-mately 35 seconds. Auditory presen tations began every60 seconds thereby allowing a 25 second pause betweeneach delivery. Three seconds before the announcer be-gan with the copy proper he said "ready" in an effort toset the subjects. This corresponded to the set produced


7/22

MEMORY FOR ADVERTISING 5 1in the visual series by the signal "tu rn ". W ith GroupsA and B the same method was followed except thatauditory presentations came first.Recall and recognition tests were given at three dif-ferent time intervals after the presentations, 1 day, 7days, and 21 days. Three subgroups of subjects wereformed for this purpose. A subgroup was made up byselecting every th ird name from the class rolls. On thetest day the instructors read the names of the studentswho were to serve as subjects. They left their classesand met in a common room. Thus each group was testedonly once. In our determinations of effectiveness of thetwo media, we were concerned only in. correct responsesto trade-name-commodity-associations built up by bothmethods of presenta tion. Fo r each subgroup th e follow-ing tests were given in the order named: Pure Recall,Aided Recall, and Recognition.

The Pure Recall Test forms were mimeographed andheaded by the following directions to the subjects:"On the lower part of this sheet will you please list as quicklyas possible all the products you can remember as having heardor read about in this experiment. List them in the order inwhich you recall them. Try to list both product and tra de name,such a s : Beechnut Chewing Gum, not ju st Chewing Gum. Ifyou can not do that, then please list either the product or trade

name in the proper column. Please tr y to do your very best.Keep your eyes on your own sheet at all times . A sample isgiven below."Fo r th e 1 day and 7 day group s approximately 20 min-utes were required before everyone finished and theblanks were collected. The 21 day groups required about10 minutes on the same form.The Aided Recall Test forms were given as soon asall Recall blanks had been collected. Two form s of theAided Recall Tests were designed in order to keep thecommodity names on the blank in the same sequence asthat in which they had been in the booklet and through


8/22

52 FRANK N . STANTONthe loudspeaker or vice versa. This test form consistedof a series of 16 blank lines with one commodity appear-ing to the right of each line. Only products actuallyused in the experiment were included in the list Theinstructions on each form of the test were as follows:

"On the lower part of this sheet you will find a list of productsmade up from the advertisements you heard and read in theexpe rimen t. In front of each product on the list is a blank space.Write in the blank space the trade name for each product onthe list th at you can remem ber. Please keep your eyes on yourown paper a t all times. A sample is given below."

Not over 10 minutes were used by any test group insatisfying the conditions of this form.The Recognition Test blanks were made up likewiseand consisted of 16 blank lines each of which was fol-lowed on the right by the product and four, two syllable,fictitious tra de names. One out of each four was correctand was to be recorded in the blank. The positions ofthe correct trade names were staggered down the page.Directions for this test follow:"On the lower part of this sheet you will find a list of productsmade up from the advertisements you heard and read in theexperiment. In fron t of each product is a blank space in which

you are to write the correct trade name of each product if yourecognize the trade name in the group of four trade names appear-ing to the rig ht of each product listed. Please keep your eyeson your own sheet at all times. A sample is given below."After the completion of the three test forms, thegroups were asked by the experimenter not to informthe remaining subjects of the nature of the testing.

Obviously the remaining subjects knew of the testingand that in all probability they too would be used.The individual test forms were scored and tables, in-cluding the means, standard deviations and analyses ofsignificant differences were built up for each group oneach of the three test periods for the three types of tests.


9/22

MEMORY FOEADVERTISING 53From TableI it canreadily beseenwe arepreparedtomake several comparisons between auditory and visualscores. Tofacilitate thevisualization of these compari-sonsor breakdowns Table II has been prepared. It isidenticaltoTableIexceptfor the connecting lines whoseencircled numbers labelthecom parisons.Inouranalysisof the data collectedin this experimentwe shall makeall the possible auditory-visual compari-sons that will contribute worthwhile information. Forinstance: Comparison 1involves different subjects,onehaving CopyXaud itoryas itsfirst presentationand theother having CopyXvisualas its first:

TABLE II

>\ U i auJ Xietnw ,


10/22

54 PRANK N . STANTON

TABLE IIIResults: 1 Day after Presen tation

flltla

V3

gO'SbeoCJK

1

iQ123456789101112123456789101112123456789101112

2

oyMe

u

2.923.614.762.712.923.614.762.712.923.612.714.763.153.695.002.863.153.695.002.863.153.692.865.007.386.776.826.297.386.776.826.297.386.776.296.82

3

S 3t ?ou

131317141313171413131417131317141313171413131417131317141313171413131417

4

bbou

1.731.982.291.831.731.982.291.831.731.981.832.292.891.732.172.382.891.732.172.382.891.732.382.17.62.961.341.71.62.961.341.71.62.961.711.34

5

IMe

3.762.282.002.382.002.383.762.282.382.003.762.284.472.282.462.692.462.694.472.282.692.464.472.287.006.775.617.005.617.007.006.777.005.617.006.77

6

SB171413131313171413131714171413131313171413131714171413131313171413131714

7

bd03

2.311.901.411.441.411.442.311.901.441.412.311.902.642.372.371.532.371.532.642.371.532.372.642.371.141.201.73.961.73.961.141.20.961.731.141.20

8 9

S

ier

.841.332.76.33.92 .521.23 .211.00 .40.43 .36.541.611.052.481.321.412.54.17.69 .601.00 .42.53 .38.58 .47.461.231.612.72.38.001.21.711.77 .69.23 .66.18 .02.48 .33.381.16.71.05

10

crdl

.74.75.67.71.43.61.60.62.62.67

.81.751.02.79.85.76.66.49.66.65.90.82.80.82.33.00.57.53.38.12.42.46.32.55.54.45

11

rd1.141.774.12.052.142.011.67.69.872.401.293.301.291.782.99.221.042.04.80.89.511.502.013.321.15.002.121.344.661.92.431.041.192.111.31.01

12 13

1.0598.30-201.7146.10-053.6977 . 01*.4984 .70-602.0004 .10-051.9156.10-051.5313.20-101.8456.10-05.8406 .50-402.3232 .05-021.3353.20-103.1302 .01*1.2579.30-201.6887.20-102.9481 .01*.2103 .90-801.3487.30-201.8416.10-05.7383 .50-40.8631 .50-40.4942 .70-601.4713.20-101.7083.10-053.2227 .01*1.0497.40-302.0865 .05-021.2681.30-204.6838.01*.9975 .40-30.4028 .70-601.0474.40-301.1654.20-102.0565 .10-051.3353.20-10.10441.0 .90

14

oSp.

VAAAAAAAAAVAVAAAAAAAAAVAA0AVAVVVAAVA

P Is less than .01


11/22

Days after Presentation


12/22

Graph 2AIDED RECALL

5

3

2

Groups

Graph 3RECOGNITION

LEG END Visual Scores

Auditory Scoreseassas ^ A verage of Visual ScoresAverage of A uditory Scores ^

Encircled X or Y with 1 or 2 denotesthe copy (X or Y Series) used and orderof presentation in the four groups

Graph 4AVERAGES

2

- 1 -

Recall

21 1 021 1 2er Presentat ion Days af ter Presentat ion Days after Presentation


13/22

MEMORY FOB ADVERTISING 5 58), the significance of these differences as determined bythe conventional critical ratio9 (Column 11), and also byusing Fisher's "t" function10 designed for small com-parisons (Columns 12 and 13 ). In view of the fact th atnone of the groups compared in Tables III, IV and V islarger than 17, more weight should be given to the find-ings based on Fishe r's formulae. The critical ratios areincluded only for purposes of comparison. Column 14denotes by either the letter A or V which medium issuperior.Graphs 1, 2 and 3 were prepared to illustrate certainof the values shown in the Tables II I, IV and V. Onedecided difference between the treatment given the tablesand the graphs should be noted. Table II I for instanceis based on the results of testing with the three types oftests only 1 day after the presentation. Graph 1, how-ever, presents a picture of the results taken not only 1day after presentation but also 7 days and 21 days, forthe Recall Tests only. Graph 2 illustrates the resultsof the Aided Recall Tests given on the three test days.The results of the Recognition Tests are shown in Graph3. In all three of these graphs the results of eachgroup on the visual and auditory presentations areshown. Auditory resu lts ar e shown by single lines andthe visual scores by double lines.In order to denote which of the two methods of presen-tation came first or second, without referring to Table I,the various lines are broken by circles which encloseeither the number 1 or 2. The letters X and Y in the cir-cles designate the copy series used. The groups are tiedtogether with lines running between the pairs of circles

o For discussion of this technique see G ar ret t, H . E., Statisticsin Psychology and Education pp. 128-134, 286-288, Longmans.Green and Co., New York, 1926.io Fo r details of sta tistica l technique of the " t" function seeFisher, R. A., Statistical Methods for Research Workers pp. 99-111, Third Edition, Oliver and Boyd, Edinburgh, 1930.


14/22

5 6 PRANK N . STANTONin order to aid in showing the group trend s. Each pa iris marked by its group letter. In addition to these lighterlines on each graph th ere are two bold lines. These rep -resent the average of all results for each medium on theparticu lar test form shown in the graph . More will besaid of the heavier lines later.An inspection of Table III shows that for pure recallone day after the presentations the auditory method issuperior in 10 out of the 12 comparisons (Table II).This decided plurality for the auditory medium isprobably significant in itself but statistically only Com-parisons 3 and 12 may be regarded as genuine differencesin the light of the critical ratio and the "t" function.The 12 comparisons on the Aided Recall Tests produceabout the same picture statistically. It is worthy of notethat throughout the experiment the results of the AidedRecall Tests follow very closely the Recall Test scores.However, when the Recognition Test scores are consid-ered, the difference between the auditory and visualmethod shrinks. At 24 hours after presentation onlyhalf of the comparisons favor audition.There may be several explanations for this change.The fact that the visual responses seem to approach theauditory in recognition may be due to the fact that 8 Isa perfect score and both media are approaching theirupper limits. Again, it is argued th at one might expectinferior visual recognition of copy presented orally. Ifthe latter be true , then the resu lts shown in Table III forauditory recognition after 24 hours should be givenmore weight. Poffenberger11 points out that recall isperhaps a better check on memory where there are noattention factors operating while to make a high recog-nition score implies higher attention value of the adver-tisement. The copy used in th is experiment employed

ii Poffenberger, Albert T., Psychology in Advertising page 192,Second Edition, McGraw-Hill, New York, 1932.


15/22

MEMORY FOR ADVERTISING 5 7

TABLE IVResults: 7 Days after Presentation4 5 (5 7 8 9 10 11 12 13 14

13o

|isgp>T3S


16/22

58 FRANK N . STANTONTABLE V

Results: 21 Days after Presen tation

aoPH

gKS

go'2bouK

1

aII2345678910

1112123456789101112123456789101112

2s&lu

1.301.381.801.501.301.381.801.501.301.381.501.801.301.382.201.571.301.382.201.571.301.381.572.205.615.235.505.575.615.235.505.575.615.235.575.50

3

S5b3

131310141313101413131410131310141313101413131410131310141313101413131410

4

bb

1.201.151.471.341.201.151.471.341.201.151.341.471.201.151.89.901.201.151.89.901.201.15.901.892.121.371.961.592.121.371.961.592.121.371.591.96

5

1.401.00.38.69.38.691.401.00.69.381.401.001.501.14.61.77.61.771.501.14.77.611.501.146.005.433.154.693.154.696.005.434.693.156.005.43

6

z3101413131313101413131014101413131313101413131014101413131313101413131014

7

b1

1.911.25.491.32.491.321.911.251.32.491.911.252.461.50.621.58.621.582.461.501.58.622.461.501.791.291.991.811.991.811.791.291.811.991.791.29

8 9 10

91 aM la O.10 .68.38 .461.42 .48.81 .52.92 .03 .35.69 .46 .36.40 .75 .40.50 .48 .35.61 .491.00 .34.10 .69.80 .58.20 .85.24 .511.59 .61.80 .50.69 .47 .29.61 .46 .41.70 .50 .70.43 .45 .36.53 .55.77 .36.07 .821.06 .70.39 .82.20 .51

2.35 .83.88 .652.46 .58 .52.54 .59 .41.50 .44 .60.14 .37 .43.92 .772.08 .67.43 .70.07 .70

11

3S.15.832.961.562.631.92

1.001.431.242.941.451.38.23.472.601.602.381.491.001.19.96

2.14.081.51.48.392.831.354.731.32.83.321.193.06.61.01

12 13

.1470 .90-80.7891 .50-403.1113 . 01*1.5212.20-102.0004 .10-053.3020 . 01*1.3094.30-201.6138.20-101.1995.30-202.8023 .01*.1439 .90-801.3755.20-10.2450 .90-80.4465 .70-60

2.7138 .02-011.5706.20-102.4073 .05-022.5804 .05-021.6560.20-101.2557.30-20.9366 .40-302.0668 .05-02.09401.0 901.4672.20-10.4465 .70-60.3764 .80-702.6996 .02-011.2941.30-205.1727 .01*1.1710.30-20.6680 .60-50.3076 .80-701.1566.30-203.0184 .01*.5937 .60-50.10081.0 90

14

oVcVAAAAAAAAAAAVAAAAAAAAAAAVVAAAAVAAAVA

P is less than .01


17/22

MEMORY FOE ADVERTISING 5 9no atten tion getting devices voluntarily. Consequently,on th at basis we would expect a shrinkage in the differ-ences between the two media.Table IV presents the data for the scores made 7 daysafter the presentation . The differences a t th is point inthe testing are the greatest of the three testing dates forthe recall tes ts. All 12 comparisons on the Recall Testscores favor the auditory method. Eleven out of the 12Aided Recall Test differences are in the direction ofaudition . Seven of the 12 auditory differences for recallare clearly significant. Nine out of the 11 of the differ-ences favoring audition on the Aided Recall Tests like-wise are statistically significant. Recognition Test scorestaken at this period show a slight superiority in favorof aud ition with 7 out of 12 in th a t direction. One ofthe 7 differences is significant according to the data.

The data secured from the tests given 21 days afterthe presentation show a decided fall in the amount re-membered according to all three test forms. Materialdelivered by the auditory method still leads on all testforms but the difference is very slight in many cases.The Recall Test results at 21 days show 11 out of 12comparisons in favor of audition. Three of the 11 differ-ences ar e statistically significant. The results of theAided Recall Tests are very close with 11 out of 12 onthe side of audition. Here one of the differences is sta-tistically significant. Auditory recognition results arestrongest at the 21 day period for 8 out of the 12 differ-ences are in that direction, three of them being definitelysignificant. The increasing difference between th e audi-tory and visual recognition results is clearly shown inGraph 3 by the constant increasing divergence betweenthe average lines for the two media of the te st periods.Up to this point the data have been presented by groupbreakdowns ju st as the material was presented. Thishas been done in order to give a fair picture of the re-sults because certain differences and significant varia-


18/22

6 0 FRANK N . STANTONtions are oftentimes obscured when scores obtained underslightly different conditions are grouped and averagescomputed. However, to get an idea of the general trendwe have grouped all the auditory scores and all the visualscores for each test on the three test days. The groupinginvolves all the auditory and visual scores of Compari-sons 5, 6, 7 and 8. Out of the 12 possible auditoryvisual comparisons (Table II) it seems reasonable thatthe comparisons on the same subjects 5, 6, 7 and 8 shouldbe given more weight. The reversa ls of the methods ofpresentation together with the changes in the order ofcopy reveal no extreme differences due to these positionalfactors. We have assumed tha t the effects of these vari-ables have been balanced out in the averages, e. g. CopyX occurred first as many times as second and the visualmethod of presentation was used first as many timesas second.The results of this grouping are shown in Table VI.In this table the data are given for the three test periodsand for the three types of tes ts. Differences have beencomputed between the auditory and visual averages foreach test form on each test period of the grouped scores.From the differences the critical ratios have been calcu-lated and are also presented in Table VI. The auditorymethod of presentation is superior on each test day forall th ree types of tes ts. Statistically, only the compari-sons between audition and vision for the Recall Tests 1day, 7 days and 21 days after presentation and the com-parisons for Aided Recall Tests 7 days and 21 days afterpresentation are significant.Graph 4 illustrates the comparisons made in TableVI. I t is identical to the bold or average lines shown inGraphs 1, 2 and 3. From this graph it is very evidentthat the responses made to the Aided Recall Tests areju st a little higher than those to the Recall Tes ts. Theauditory trend is indicated by solid black and the visualby outline. The grea tes t difference between the aud itory


19/22

MEMORY FORADVERT ISINGT A B L E VI

61

1

1

R

ddR

R

o

MeanaDifferencer


20/22

62 FRANK N . STANTONthe auditory recall curve is noted while the visual trendis also parallel to the curve shown in Graph 4 of thisreport.The approximate identity of the Recall and Aided Re-call curves can also be clearly seen in G raph 4. Recog-nition for the two methods of presentation seems to bevery nearly equal after a lapse of one day, but as thetime goes on the auditory method increases its lead overthe visual technique. That is con trary to the resu ltsobtained for the Recall and Aided Recall testings for inboth the peak difference occurred at 7 days. I t shouldbe noted th a t we obtained d ata at only 3 poin ts. Conse-quently, it is quite possible tha t th e actual peak differ-ence might occur somewhat earlier or later than 7 days.The results of the experiment agree within limits withthe work done by DeWick. He employed only a recalltes t which was given to all subjects immediately, 1 day, 7,and 150 days after stimulation. Immediate recall showedvery little difference in favor of the auditory medium.This difference increased up to 5 days in Series I andup to 7 days in Series II.Chiefly because the results would have few practicalapplications, and secondly due to the equality of resultsDeWick secured, the immediate test period was omittedin the present study. The 1 day, 7 day and 21 dayperiods were selected arbitrarily.An examination of Graph 4 will give some indicationof the magnitude of the correct responses for each testform. Auditory recognition scores ru n well ahead of theAided Recall and Recall mean scores. Such resu lts arenot con trary to the usual comparisons of this na ture . Ofthe three test forms, perhaps the Aided Recall test isnearest to answering questions of a practical nature thatare raised by the advertiser when considering which ofthe two media he will select to carry his story. If heis advertising products similar to the ones used in thisstudy, his chief desire is to have the consumer think of


21/22

MEMORY FOE ADVERTISING 6 3his brand or trade name when a need arises for thatproduct.Throughout the experiment every precaution wastaken to control all the variables by keeping them asconstant as possible. Yet in searching the procedure andresults for an answer to the reason for auditory supe-riority, one must not overlook the part played by theexperimental situation. The auditory results are prob-ably heightened somewhat as a result of the abnormalatten tion to the loudspeaker and the copy. But one mustnot overlook the visual advantages. Few prin ted adver-tisements receive the attention given to the visual copyin this study. The students were warned not to lookinside the booklets, thus setting up a condition of ex-pectancy. They were instructed to read every word ofcopy which, of course, is not the case in normal visualadvertising. Thus the effectiveness of both media pre-sumably was stepped up . Here, as in the case of pre-senting the copy (visual void of illustrations, displayfaces, etc.auditory without music or sound effects), theadvantages and disadvantages are as nearly equal as pos-sible with this set-up. We must not overlook the par-ticipants who were college students, and their surround-ings, hab its, ages and intere sts. The w riter plans tocontinue this means of measurement on listeners in homesituations . When we consider th at the college studentis a trained reader, such an experiment with other per-sons may even show a greater difference in favor of audi-tion for certain economic levels.

SUMMARYFictitious advertising copy was presented visually bymeans of printed booklets and orally through a smallpublic address system to four groups of college students.Testing 1 day, 7 days and 21 days after stimulation byRecall, Aided Recall and Recognition tests for correct


22/22

6 4 PRANK N . STANTONtrade-name-commodity-association built up by the twomedia showed the aud itory method superior. Not allcases showing an auditory superiority were statisticallysignificant. The peak of the superiority in the RecallTests came for the 7 day period. A t the 21 day checkboth auditory and visual results were much lower andthe margin between the two had suffered by the lapse oftime. Throughout the experiment little difference inscores was noted between the Aided Recall and Pure Re-call tests . In the recognition scores, the greatest audi-tory superiority occurred at 21 days. The principalvariable that might indicate the desirability of qualify-ing the results was the discrepancy arising from theexperimental situation between actual reading-listeningconditions and the ones set up in the study.

advertising visual vs auditive

Documents