UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Proposal for audience measurement in print media Jones, Vernon J. 1970

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1970_A4_5 J65.pdf [ 3.84MB ]
Metadata
JSON: 831-1.0101974.json
JSON-LD: 831-1.0101974-ld.json
RDF/XML (Pretty): 831-1.0101974-rdf.xml
RDF/JSON: 831-1.0101974-rdf.json
Turtle: 831-1.0101974-turtle.txt
N-Triples: 831-1.0101974-rdf-ntriples.txt
Original Record: 831-1.0101974-source.json
Full Text
831-1.0101974-fulltext.txt
Citation
831-1.0101974.ris

Full Text

A PROPOSAL FOR AUDIENCE MEASUREMENT IN PRINT MEDIA by VERNON J. JONES B.A., University of B r i t i s h Columbia, 1968 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF BUSINESS ADMINISTRATION in the Faculty of Commerce and Business Administration We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH October, 1970 COLUMBIA In presenting th i s thes i s in pa r t i a l fu l f i lment of the requirements for an advanced degree at the Un ivers i ty of B r i t i s h Columbia, I agree that the L ibrary shal l make i t f ree l y ava i l ab le for reference and study. I fur ther agree that permission for extensive copying of th i s thes i s for scho lar ly purposes may be granted by the Head of my Department or by his representat ives. It is understood that copying or pub l i ca t ion of th is thes is fo r f inanc ia l gain sha l l not be allowed without my wr i t ten permission. Department of Commerce and B u s i n e s s A d m i n i s t r a t i o n The Univers i ty of B r i t i s h Columbia Vancouver 8, Canada D a t e Qctnbpr 6, 1970 ABSTRACT A major concern among advertisers and media managers i s the measurement of net audience coverage achieved by an advertising campaign over time and across combinations of publications. Measures of audience exposure for combinations of publications have been shown to be more accurate when based on audience segments associated with each publication than when based on aggregate exposure to a l l the publications i n the group. This thesis argues that the concept of duplication among audience segments associated with a combination of i n d i v i -dual publications i s equally applicable to the segments associa-ted with the sections of a single publication. Accordingly, i t i s the objective of t h i s thesis to demonstrate that audience measures based on audience segments associated with sections of a publication are superior to those measures based on aggregate exposure to that publication. The fundamental measures of audience exposure are un-duplicated audience or net reach, duplicated audience and average frequency of exposure. The relationships among these measures were developed i n a t h e o r e t i c a l model of i n t e r -section duplication. The model was then applied to data drawn from a recent study on a major Canadian newspaper. As any application of the segmented audience concept depends on a simple and accurate method of estimating net reach for a c o m b i n a t i o n o f s e c t i o n s , c o n s i d e r a b l e e f f o r t was expended t o d e s c r i b e r e c e n t r e s e a r c h c o n c e r n i n g e s t i m a t i o n of n e t r e a c h f o r c o m b i n a t i o n s o f p u b l i c a t i o n s and t o r e l a t e such r e s e a r c h t o the o b j e c t i v e s o f t h i s t h e s i s . I t was c o n c l u d e d t h a t segmented audience d a t a are s u p e r i o r t o aggregate d a t a as a b a s i s f o r audience measure-ment, and t h e r e f o r e , an a d v e r t i s e r must e v a l u a t e , a c c o r d i n g t o a d v e r t i s i n g o b j e c t i v e s , the placement o f h i s a d v e r t i s e m e n t s and the i n h e r e n t t r a d e - o f f between net r e a c h and f r e q u e n c y f o r a g i v e n a d v e r t i s i n g campaign. The paper c l o s e s w i t h some s u g g e s t i o n s f o r f u r t h e r s t u d y . TABLE -OF CONTENTS CHAPTER Page I. INTRODUCTION 1 I I . RESEARCH SETT ING . . . . . . . . . . . . . . . . . . . . . . . . 6 I I I . RELATED L ITERATURE 9 IV. HYPOTHESES AND THEORETICAL MODEL 25 V. METHOD OF ANALYS IS 38 V I . SUMMARY OF RESULTS 47 V I I . CONCLUSIONS 66 BIBLIOGRAPHY • 75 APPENDICES . . . 76 i i LIST OF TABLES TABLE Page I. Net Reach as a Percentage of Campaign Size at Various Levels of Readership 51 II . Comparison of Exposure Results of Six Advertising Campaigns and Aggregate Newspaper Exposure . 5 6 I I I . Comparison of Exposure for Newspaper Sections and Aggregate Newspaper Exposure 57 IV. Comparison of Exposure to Five Campaigns: 1 through 3 Placed Exclusively i n one Section, 4, 5 Placed across Sections 58 V. Comparison of Exposure for Five Campaigns According to Sports, Finance, Women and Entertainment Sections . . 60 VI. Calculation of DA, UDA and F for Campaign 4 . 61 VII. DA, UDA, F for Campaign Alternatives 1 through 16 64 VIII. DA, UDA, F for Campaign Alternatives 1 through 16 Arranged According to Increasing Net Reach 65 i i i LIST OF FIGURES FIGURE Page 1 Audience Duplicated by Two Vehicles 11 2 Observed Relationship Between Unduplicated (z) and Pair-wise Duplicated (x) C o e f f i c i e n t s of the Total Audience 13 3 Model of Intersecting Audience Segments 31 4 Individual Section Readership 40 5 Multiple Section Readership 41 6 Representation of Data Block: Individual Sections 4 3 7 Representation of Data Block: Paired Sections . 44 8 Section/Week Decision Matrix: Panels A, B , C Combined . . . . 48 9 Section/Week Decision Matrix: Panel A 49 IV ACKNOWLEDGEMENT I am greatly indebted to Professor F. H. S i l l e r for his d i r e c t i o n and invaluable assistance i n the pre-paration of this thesis. CHAPTER I INTRODUCTION Purpose The g e n e r a l o b j e c t i v e o f t h i s t h e s i s i s t o demonstrate the inadequacy o f u n d u p l i c a t e d r e a c h and f r e q u e n c y o f exposure c a l c u l a t i o n s f o r p r i n t media, when t h e s e measures are based on a g g r e g a t e audience d a t a . * T h i s t h e s i s w i l l show the supe-r i o r i t y o f measures based on segmented audience d a t a w h i l e a t t h e same time acknowledging the l i m i t a t i o n s o f such measures when a d v e r t i s i n g campaigns c u t a c r o s s s e v e r a l audience segments. F i n a l l y , i t w i l l be i l l u s t r a t e d t h a t , by d e f i n i t i o n , undup-l i c a t e d r e a c h and fr e q u e n c y of exposure measures a r e i n v e r s e l y r e l a t e d and hence cannot be s i m u l t a n e o u s l y maximized. The l a t t e r i s an i m p o r t a n t b u t o f t e n o v e r l o o k e d c o n s i d e r a t i o n i n the p l a n n i n g o f campaign s t r a t e g i e s . Importance o f the Study I n e v a l u a t i n g the r e a d e r s h i p l e v e l o f p r i n t media the fundamental measures o f performance a r e : the u n d u p l i c a t e d a u d i e n c e o r n e t r e a c h , the d u p l i c a t e d audience and the average f r e q u e n c y o f exposure. However, t h e r e has been some j u s t i f i e d * The r e a d e r who i s u n f a m i l i a r w i t h the t e r m i n o l o g y of audien c e r e s e a r c h may f i n d a complete d e f i n i t i o n o f terms used i n t h i s t h e s i s b e g i n n i n g on page 27. 2 . m i s g i v i n g s about t h e c a l c u l a t i o n o f t h e s e measures and the uses t o wh i c h they have been p u t . Dr. L. B o g a r t of the American Newspaper P u b l i s h e r s A s s o c i a t i o n wrote an a r t i c l e i n 1966, p u b l i s h e d i n the J o u r n a l o f M a r k e t i n g , s e v e r e l y c r i t i c i z i n g a udience measurement t e c h n i q u e s . 1 He s t a t e s t h a t t o t a l a udience f i g u r e s do not t r u l y r e p r e s e n t the market and p r o v i d e m i s l e a d i n g c r i t e r i a f o r a d v e r t i s i n g d e c i s i o n -making. He goes on t o s t a t e t h a t r e s e a r c h e f f o r t would be b e t t e r c o n c e n t r a t e d i n communications and l e a r n i n g t h e o r y but he a l s o r e c o g n i z e s the v a l u e o f s t u d i e s w h i c h attempt t o d e f i n e a u d i e n c e s i n more d i f f e r e n t i a t e d terms than i s done p r e s e n t l y . He c i t e s as an example t h e s t u d i e s done by A l f r e d P o l i t z R esearch. There has, i n f a c t , been c o n s i d e r a b l e r e s e a r c h i n r e c e n t y e a r s t o make audience measures more r e a l i s t i c by ex-p e r i m e n t i n g w i t h t h e d a t a p r o v i d e d by A l f r e d P o l i t z Research and s i m i l a r s t u d i e s i n Europe. A major e f f o r t has been made t o r e l a t e u n d u p l i c a t e d audience and f r e q u e n c y o f exposure v a r i a b l e s w i t h s i m p l i f i e d a n d . o p e r a t i o n a l c o n c e p t s . Most of the r e s e a r c h i n p r i n t media has been c o n c e n t r a t e d on magazines. T h i s t h e s i s i s an e x t e n s i o n o f t h a t r e s e a r c h . G e n e r a l l y , the importance o f t h i s t h e s i s i s t o f a c i l i -t a t e more e f f i c i e n t a d v e r t i s i n g d e c i s i o n s by p r o v i d i n g more r e f i n e d measures than the c l a s s i c i a l aggregate i n d i c a t o r s such as c i r c u l a t i o n and d u p l i c a t e audience d a t a . More s p e c i f i c a l l y , 3. i t i s intended to relate recent research concerning duplication among magazine audiences to duplication among section audiences within an i n d i v i d u a l publication, i n thi s case a newspaper. To the advertising decision-maker, th i s study w i l l demonstrate that aggregate audience measures may not r e f l e c t the true nature of an audience's c h a r a c t e r i s t i c s and w i l l show that there i s a trade-off between net reach and frequency of exposure for a campaign i n an i n d i v i d u a l publication. Further, i t w i l l point out that to receive the highest return per advertising d o l l a r , the decision-maker must evaluate t h i s trade-off according to his advertising objectives. Scope and Limitations This paper w i l l make extensive use of data collected during a recent newspaper readership study. Although these data provide a detailed record of longitudinal readership p r o f i l e s , certain l i m i t a t i o n s should nevertheless be ci t e d . F i r s t , the o r i g i n a l study considered only a single newspaper over a. l i m i t e d period of time and was carr i e d out only i n a c i t y having one major newspaper. As a r e s u l t , although t h i s thesis i s intended to be c h a r a c t e r i s t i c of a l l forms of p r i n t media, such a generalization may be distorted due to the unique circumstances of the o r i g i n a l investigation. Second, the o r i g i n a l study was not intended to te s t the s p e c i f i c hypotheses of this thesis, and, as a r e s u l t , there are some rather serious gaps i n the data that necessitate a reduction 4. i n the sample si z e . This problem w i l l be discussed more f u l l y i n Chapter VI.* Plan of the Study Chapter II gives a detailed description of the study from which the data for this paper were drawn. I t considers the objective of that project and the method of data c o l l e c -tion including a description of the sample and the question-naire. Chapter III surveys the l i t e r a t u r e related to the objective of the thesis. I t i s primarily concerned with net audience estimation and draws on research originated by J. M. Agostini concerning French magazines. Chapter IV states and discusses the research hypotheses. The indices of audi-ence measurement, reach and frequency, are defined, as well as other variables fundamental to the hypotheses. These variables are then related i n a t h e o r e t i c a l model to demon-strate the l o g i c of the research procedure and the generation of the hypotheses. Chapter V describes the method by which the available data were organized and then analyzed to test the hypotheses of Chapter IV. The method of analysis i s the detailed exposition of the t h e o r e t i c a l model applied to the data. Chapter VI considers d i f f i c u l t i e s i n data analysis and presents the research findings with explanation and some * This problem i s referred to again i n Chapter I I . Dis-cussion i s deferred to Chapter VI as f a m i l i a r i t y with Method of Analysis (Ch. V) i s necessary before the problem can be c l a r i f i e d . 5. discussion of those findings. F i n a l l y , Chapter VII presents conclusions derived from the results as well as from the chapter on related l i t e r a t u r e . This chapter also discusses implications for the advertising decision-maker and areas for further study. CHAPTER II RESEARCH SETTING The research proposal and data base for this paper were derived from one of a series of ongoing studies being conducted by a newspapers' association. The o v e r a l l project i s intended to study reach, frequency and advertising effec-tiveness of d a i l y newspapers i n Canada. The data used i n this paper are those co l l e c t e d i n the f a l l of 1968 i n co-operation with the leading newspaper of a major eastern c i t y . In response to competition from other media, newspapers have increasingly been c o l l e c t i n g audience data and developing quantitative advertising effectiveness indicators. The fundamental unit of measurement i s "opportunity for exposure," a standard made purposely analogous to those measures used by the broadcast media. For instance, i f an in d i v i d u a l has his t e l e v i s i o n turned on, he i s counted as having had an oppor-tunity to be exposed to commercial messages i n a given time period. S i m i l a r l y , i f an i n d i v i d u a l records having seen a newspaper or part of one, he has had an opportunity to be exposed to i t s advertising contents. Three levels of"oppor-tunity for exposure" were designated i n the study: exposure to the entire newspaper, exposure to the in d i v i d u a l pages and exposure to in d i v i d u a l quarter pages. 7. Method o f Data C o l l e c t i o n Each F r i d a y morning the respond e n t s r e c e i v e d a s e l f -a d m i n i s t e r e d q u e s t i o n n a i r e which was c o l l e c t e d l a t e r the same day. The q u e s t i o n n a i r e i n c l u d e d an a b r i d g e d m i n i a t u r i -zed v e r s i o n of t h e p r e v i o u s n i g h t ' s newspaper. Each page o f the m i n i a t u r i z e d paper was d i v i d e d i n t o f o u r q u a r t e r s and the respondent r e c o r d e d whether o r n o t he had seen the paper, seen p a r t i c u l a r pages of i t and seen p a r t i c u l a r q u a r t e r pages. Due t o the d i f f e r e n t o b j e c t i v e s of the so u r c e s t u d y and l i m i -t a t i o n s on r e s o u r c e s , the q u e s t i o n n a i r e r e p r o d u c t i o n was a b r i d g e d . Not eve r y page i n the p r e v i o u s n i g h t ' s paper was i n c l u d e d i n t h e q u e s t i o n n a i r e . The r e s u l t a n t d i f f i c u l t i e s a r e d i s c u s s e d f u l l y i n Chapter V I . The m i n i a t u r i z e d newspaper was one among o t h e r media q u e s t i o n n a i r e s d i s t r i b u t e d once a week f o r s i x weeks. The re a s o n f o r i n c l u d i n g o t h e r media q u e s t i o n n a i r e s was t o p r e v e n t the respond e n t s becoming s e n s i -t i z e d t o the f a c t t h a t they were p a r t o f a newspaper s t u d y . The r e s p o n d e n t s t o the q u e s t i o n n a i r e f i t t e d i n t o one of t h r e e p a n e l s o r g a n i z e d on a g e o g r a p h i c a l b a s i s . * The c i t y was d i v i d e d i n t o two s e c t i o n s . The re s p o n d e n t s i n c l u d e d i n p a n e l A were drawn e n t i r e l y from one o f t h e s e g e o g r a p h i c a l s e c t i o n s , t h e re s p o n d e n t s f o r p a n e l B e n t i r e l y from t h e o t h e r * The g e o g r a p h i c a l areas were matched as c l o s e l y as p o s s i b l e a c c o r d i n g t o demographic c h a r a c t e r i s t i c s . 8. s e c t i o n . Both groups were s e l e c t e d randomly w i t h i n the g e o g r a p h i c a l l i m i t a t i o n s . P a n e l C was s e l e c t e d randomly from the e n t i r e c i t y . The r e s u l t i n g sample s i z e was 1,220 r e s p o n d e n t s : 402 i n p a n e l A, 404 i n p a n e l B and 414 i n p a n e l C. The purpose of h a v i n g t h r e e groups of res p o n d e n t s was t o t e s t the e f f e c t o f v a r i a t i o n s i n a d v e r t i s e m e n t s . The e f f e c t i v e n e s s of c e r t a i n a d v e r t i s i n g campaigns c o u l d be t e s t e d by v a r y i n g t h e i r placement and c o n t e n t . Hence p a n e l s A, B and C would r e c e i v e s l i g h t l y d i f f e r e n t newspapers and t h e q u e s t i o n n a i r e s d i s t r i b u t e d would a l s o v a r y . T h i s p r o c e d u r e has no r e l e v a n c e f o r t h e s i s e x c e p t f o r the t a b u l a t i o n problems i t c r e a t e s . T h i s paper i s not concerned w i t h the e f f e c t i v e -ness of campaign c o n t e n t and has no r e a s o n t o i n t e n t i o n a l l y d i s t i n g u i s h among any of the 1,220 r e s p o n d e n t s . However, such a d i s t i n c t i o n becomes n e c e s s a r y and a g a i n the r e a d e r i s r e f e r r e d t o C h a p t e r VI f o r f u r t h e r d i s c u s s i o n o f the problem. The exposure r e s u l t s f o r the sample of r e s p o n d e n t s o v e r the s i x weeks r e p r e s e n t the d a t a base f o r t h i s t h e s i s . From t h e s e , the measures of u n d u p l i c a t e d and d u p l i c a t e d a u d i e n c e s and f r e q u e n c y o f exposure f o r v a r i o u s c o m b i n a t i o n s o f s e c t i o n s o f the newspaper w i l l be d e r i v e d . CHAPTER III RELATED LITERATURE Considerable research has been conducted i n the past few years concerning audience measurement i n the p r i n t media. The primary objective has been to f i n d a simple method for estimating the unduplicated audience of a combination of magazines. I t has been known for some time that when an advertiser places a campaign i n a number of publications, the exposure of the campaign w i l l be duplicated due to the fact that there i s overlapping readership of those publications. The problem t h i s has created i s to distinguish the net undup-l i c a t e d reach of the campaign without the very extensive and costly tabulation of a l l the d i f f e r e n t possible combinations of p r i n t media readership. I t i s worth noting one or two examples which demonstrate the i m p o s s i b i l i t y of actually c a l -culating a l l the available p r i n t combinations. Referring to a German magazine readership study conducted i n 1961, Walther Kuhn estimated that for the 44 magazines used i n the study 2 there were 17.6 b i l l i o n combinations. In a similar study i n France by the Centre d'Etude des Supports de P u b l i c i t e 3 (CESP), on which the o r i g i n a l research discussed i n this chapter was based, the 30 magazines provided 1,073,741,793 * combinations. Even computer assisted, the achievement of * The comparable figure for data used i n this study (see Chapter VI), though severely limited, i s 38,960 possible combinations. 10. complete r e s u l t s f o r l a r g e s t u d i e s can be presumed u n a t t a i n -a b l e . Most o f the r e s e a r c h conducted i n t h i s a r e a has depended upon th e development of an e m p i r i c a l r e l a t i o n s h i p between the u n d u p l i c a t e d a u d i e n c e of a c o m b i n a t i o n o f magazines and the d u p l i c a t e d a u d i e n c e between p a i r s o f magazines i n c l u d e d i n the c o m b i n a t i o n . J . M. A g o s t i n i i n the J o u r n a l o f Adver-t i s i n g R e s e a r c h , March, 1961, was the f i r s t t o p r e s e n t t h i s 4 e s t i m a t i n g p r o c e d u r e . U s i n g a c t u a l measures of u n d u p l i c a t e d r e a c h and d u p l i c a t e d r e a c h f o r a l l p o s s i b l e c o m b i n a t i o n s of 15 of t h e 30 magazines used i n the CESP st u d y (32,767)combina-t i o n s ) , A g o s t i n i d e v e l o p e d e m p i r i c a l l y a f o r m u l a f o r the s i m p l e e s t i m a t i o n o f u n d u p l i c a t e d a u d i e n c e f o r any o f the 30 magazines and t h e n t e s t e d i t s g e n e r a l i t y on a s t u d y of l e a d i n g American magazines. A g o s t i n i ' s method and r e s u l t s d e s e r v e c a r e f u l c o n s i d e r a t i o n as they a r e fundamental t o s u c c e e d i n g l i t e r a t u r e d i s c u s s e d i n t h i s c h a p t e r and p r o v i d e i n t e r e s t i n g i m p l i c a t i o n s f o r the f u r t h e r development o f the r e s u l t s t o t h i s t h e s i s . L e t a,b,c,...,n r e p r e s e n t the d i f f e r e n t media o r * v e h i c l e s i n a c o m b i n a t i o n and A , A, ,..., A t h e i r r e s p e c t i v e a b' ' n ^ a u d i e n c e s . A g o s t i n i uses the word " v e h i c l e " t o r e p r e s e n t the u n i t i n w h i c h an ad can be p l a c e d — f o r h i s s t u d y , each magazine r e p r e s e n t s a " v e h i c l e " — l a t e r , t h i s paper w i l l r e f e r t o news-paper s e c t i o n s as v e h i c l e s . (1) A A + A, + a b + A n Let C equal the net coverage of the combination of media. Because of the duplication among readers of the media C i s usually smaller than A. ( 2 ) C = zA where 0 < z < 1 The method depends upon the c a l c u l a t i o n of z from a matrix showing the duplicated audience of a l l the possible pairs of vehicles i n the media combination. D, the sum of these duplications, i s the half-sum of the terms i n the matrix, as i t i s symmetrical about the main diagonal.(Fig. 1). Figure 1 Audience Duplication by Two Vehicles n n D , D ab ac ba D D , ca cb D be D D , D na nb nc an D bn . D cn Let x = D/A In order to generate x, then the sum of the media audiences (A) and the sum of the pair-wise duplications (D) must be known. Agostini's objective was to demonstrate a relati o n s h i p between x and z.. Obviously the greater the duplication between media the less w i l l be the unduplicated audience. The higher the value of D the lower the value of C. Inasmuch as x = D/A and z = C/A, when x increases, z decreases. When there i s no duplication between media the unduplicated audience equals the sum of the audiences. In this case, when D - 0, x = 0, and when C = A, z = 1. Therefore when x = 0, z =1.6 From the available pairs of combinations Agostini selected 9 8 at random and p l o t t i n g x against z was able to describe the relationship with a continuous curve (Fig. 2). Actual co-ordinates of x and z were always within 2 percent of the curve. Deducing that the curve was . asymptotic to th x-axis, the next step was to develop an equation to f i t the rela t i o n s h i p : (3) z = f(x) (4) 1 z Kx + 1 The curve was found to be accurate where the constant K, was equal to 1.125. Figure 2 x 14. Having developed his equation, Agostini recognized i t may not be useful for data other than the CESP study. Accordingly he used i t i n further tests of other French magazines finding his estimates of unduplicated audience using K = 1.125 were at the most 2.6 percent o f f the actual figure. Also, applying the technique to a study of f i v e American magazines, i t was found that i n 96 testable combina-tions, 82 were within 1 percent error, 10 cases were between 1 and 2 percent error, and 4 cases were greater than 2 percent, the maximum being 2.7 percent. Hence, Agostini concludes that the estimate of unduplicated audience can be greatly simpli-f i e d i f one knows the t o t a l audience of the media and the two-by-two combinations. The significance for th i s thesis of the above research, as well as the succeeding research done i n the 'Agostini t r a d i t i o n , ' i s that the model Agostini developed using several publications as vehicles may.be equally applicable using in t r a - p u b l i c a t i o n sections as vehicles. In fa c t , i f i t i s presumed that people read s e l e c t i v e l y within a publication and an audience segment can be i d e n t i f i e d with each 'section' of the publication, then sections would appear to be the exact p a r a l l e l of Agostini's vehicles. Before any further discus-sion of th i s point, however, the generality of the Agostini equation should be examined. 15. The reaction to the publication was a series of attempts to test the a p p l i c a b i l i t y of both the formula and the constant, K. Although well-accepted, there were some general reservations about the Agostini equation. These were perhaps best summarized by John Bower i n J.A.R., March 1963. The Agostini method i s not the only way to estimate unduplicated audiences, but i t i s prob-ably the least complicated. I t i s based on the assumption of random readership among homo-geneous groups of vehicles. I t was constructed and tested on such combinations, i . e . , vehicles appealing to the same audience such as general c i r c u l a t i o n magazines or d a i l y newspapers. There i s thus some reason to believe that the technique may not be accurate when applied to combinations of non-homogeneous vehicles. Another possible source of inaccuracy i s the r e l a t i o n between the number of vehicles i n the combination and the error i n the estimate. The larger the number of vehicles, the more tenuous becomes any rela t i o n s h i p between duplications and net audiences.^ Using net audiences of American and Canadian magazines Bower conducted tests on seven d i f f e r e n t studies t o t a l l i n g 6 4 0 combinations using the Agostini formula. The weighted average error for the 6 4 0 cases was only 3.1 percent. Generally then he accepts the Agostini technique. The re-sults however to some extent bear out Bower's reservations as outlined above. He did f i n d that error was greater for combinations of heterogeneous vehicles, i . e . , vehicles appealing to widely varying and hence less duplicated audi-ences. He admits that he could not prove t h i s conclusively 16. within the l i m i t a t i o n s of his data. This may not be of great importance to any conclusions drawn from this thesis. The examples Bower uses are of an extremely heterogeneous character, one i n fact involving a language b a r r i e r . I t can probably be safely assumed that the homogeneity among readers of a newspaper would remove th i s e f f e c t from any application of the Agostini formula. Some of Bower's other results are more in t e r e s t i n g . He found that the error increased substantially with the number of vehicles used i n the combination. Perhaps more s i g n i f i c a n t , Bower found that although the o v e r a l l average error was small the formula overestimated the actual z/x relationship i n 90 percent of the cases, suggesting that K = 1.125 i s too small. Bower's a r t i c l e i s a useful c r i t i c i s m of the Agostini technique. However he i s forced to admit that he was not able to carry his research to the extent of providing the approp-r i a t e modifications. In the same issue of the Journal of  Advertising Research, Marcel Marc, a colleague of Agostini, demonstrated a p o s i t i v e modification to the o r i g i n a l formula under circumstances which are to some extent s i m i l a r to those 9 exi s t i n g within a publication. Marc bases his a r t i c l e on the fact that readers of the s p e c i a l i z e d press seem to be more highly duplicated than with mass magazines. He notes that Agostini's formula was derived empirically and hence should not necessarily be applied to magazines with r a d i c a l l y d i f f e r e n t types of audiences such as that associated with business or trade publications where reading i s necessary rather than entertaining. On th i s point i t appears that Marc supports Bower's hypothesis concerning the larger error where heterogeneous combinations are tested. Marc, however, carries the analysis further. While f u l l y accepting the rel a t i o n s h i p which Agostini devel-oped between x and z he repeats the analysis for a group of trade magazines testing for a more appropriate value for K. The r e s u l t was that the exact relationship between x and z were found as that for Agostini's data. However, the approp-r i a t e value of K was found to be 1.6 rather than 1.125. To summarize Marc without a detailed report of his analysis, he accepts completely the Agostini technique while forwarding a method for modifying K using a frequency d i s t r i b u t i o n of the population according to the number of magazines read. The significance of .Marc's research i s that while recognizing much the same d i f f i c u l t i e s with Agostini's fo r -mula as Bower, he concludes not that K i s incorrect at a value of 1.125 but that i t may vary under certain circumstances. He works primarily with the varying c h a r a c t e r i s t i c s of the media audience. At almost the same time J.. M. Caffyn and M. Sagovsky, working with B r i t i s h newspapers, came to the same conclusion. They concur that K may vary but on the basis of the number of publications rather than audience charac-t e r i s t i c s , which was of course the other observation of 18. John Bower. However t h e i r analysis seems weaker than that of Marc's. They do show f a i r l y conclusively that K needs modification according to the changing value of x. However, they at t r i b u t e the changing value of x to number of p u b l i -cations, as stated above, where i t seems an equally l o g i c a l explanation of the increasing r a t i o of D to A may be increased duplication of the same combination of media, i n other words a highly homogeneous mixture of publications. This would have lead them more completely along the path pursued by Marcel Marc. Walther Kuhn pursues more conclusively the e f f e c t of number of publications considered i n a combination."'""'" On tests of the Agostini formula on German magazines errors of up to 5 percent were reported. Re-estimating K, Kuhn found that a value of 1.162 gave a more accurate r e s u l t . At t h i s point Kuhn might well have written a conclusion similar to that of Marc, Bower or Caffyn and Sagovsky. Instead he went on to question the fundamental hyperbolic function independent of the value of K. From the German data, he developed as an -K v alternative an exponential function z = e n . I t was of course derived empirically and f i t t e d the data very well with less than 1.8 percent error. However no further tests were undertaken, and no further development of t h i s a l t e r -native relationship appears to have taken place. Kuhn's formula then seems to be a dead end but i s important as he 19. demonstrates f a i r l y thoroughly that the simple relationship between net audience and pair-wise duplication may well be affected by the number of publications considered in.,the set of combinations. Before discussing further research into Agostini's function and the problems of K, the role of t h i s l i t e r a t u r e should be considered within the context of the objectives of this thesis. It may prove disappointing to the reader that the hypotheses discussed i n the next chapter make no mention of K and no attempt i s made to test the " e q u i l a t e r a l hyper-bola-asymptotic to the x-axis," to quote Agos t i n i , using the data of this thesis. In f a c t , the following chapters are probably more analogous to the CESP study than to the tech-niques described i n the above l i t e r a t u r e . The reason may already be clear. The'Agostini research' i s based on the already acquired knowledge of duplication among several publications. This thesis, as already noted, i s attempting to e stablish the existence of a similar phenomenon within i n d i v i d u a l publications using newspaper sections as 'vehicles.' Limitations of data and time prevent further exploration of a relationship between net audience coverage and pair-wise duplication of pairs of newspaper sections. Hence the l i t e r a -ture discussed i n this chapter refers more to the potential of the proof of the hypotheses than to t h e i r j u s t i f i c a t i o n i n terms of present measurement techniques. This l a t t e r 20. aspect of the thesis i s expected to be drawn from the l o g i c of the t h e o r e t i c a l model discussed i n Chapter IV along with the method of analysis and would probably sound redundant i f discussed i n terms of other readership studies. Simply, then, the demonstration of a simple estimating procedure for net unduplicated audience among d i f f e r e n t magazines opens an area for speculation concerning i n t r a - p u b l i c a t i o n undup-l i c a t e d audience. Further discussion of more complicated research con-cerning Agostini's formula and the d e f i n i t i o n of K, then has an increasingly tenuous l i n k to the purpose of the thesis. However i t i s worth reviewing, perhaps i n less d e t a i l , i f only to show that the Agostini t r a d i t i o n has not been refuted. Two major contributions were made, by R. A. Metheringham and P. Hoffmans. Metheringham demonstrated a method for estimating net cumulative coverage and the frequency d i s -t r i b u t i o n of a p r i n t schedule from pairs of publications and 12 pairs of issues. His method i s t h e o r e t i c a l and uses a numerical example. The a r t i c l e i s i n t e r e s t i n g i n r e l a t i o n to t h i s thesis as he considers duplication within a p u b l i -cation. However,.it i s only the duplication a r i s i n g across successive issues of a publication with which he i s concerned whereas this thesis also considers duplication across sections within the publication. To discuss i n d e t a i l Metheringham's method would provide an unnecessary complication to t h i s chapter. The method i s s t i l l a v a r i a t i o n on Agostini. However, as an inte r e s t i n g by-product of his study, he was able to conclude that the error i n estimating K i s actually reduced with the numberoof media used i n the combination. Pierre Hofmans re - i t e r a t e d the a b i l i t y of the Agostini approach to provide accurate estimates of unduplicated audi-13 ences. Although he c a l l s for modification of the equation, he concludes that i t i s an accurate method for gathering immediate information. Hofmans, l i k e Metheringham, i s con-cerned about duplication across successive issues of p u b l i -cations. He uses the Agostini equation to develop the net reach for advertising schedules involving multiple insertions i n several media. Much of Metheringham's and Hofmans1 work was included i n an a r t i c l e by H. J. Claycamp and C. W. McClelland i n 1968 which attempts to r a t i o n a l i z e and consolidate the state of 14 knowledge concerning K. F i r s t , Claycamp and McClelland define K. They conclude i s i s a parameter, not a. constant, describing how mean and variance of a readership d i s t r i b u t i o n are related. In t h e i r analysis of data, they use a rather small readership study but their technique i s to examine the e f f e c t of alternative d e f i n i t i o n s of readership on K. Their research includes not only alternative combination of p u b l i -cations but combinations over several issues. Their readership l e v e l i s calculated on the basis of percentage of available issues read. The authors found that as the i r d e f i n i t i o n of readership was relaxed the estimated value of K increased with increases i n D/A and C/A. Further examination led them to observe that C/A was underestimated for small and overestimated for larger values of D/A. F i n a l l y , K was shown to be highly sensitive to D/A when there are few media i n the combination, but as the number of pu b l i -cations and D/A became larger the variance i n K decreases. The authors then ask, according to the i r own r e s u l t s , "Why i s i t possible to ignore the v a r i a t i o n i n K and s t i l l obtain close estimates of net coverage...?" By cal c u l a t i n g the p a r t i a l derivate of C/A with respect to K, they state the e f f e c t of errors of K on net coverage can be determined. Their r e s u l t s show that errors i n specifying K w i l l always r e s u l t i n a less than proportionate error i n C/A. Claycamp and McClelland go on to consider problems of cumulative audi-ence which w i l l not be discussed here. I t i s , however, worth noting the summary to t h e i r a r t i c l e as i t stands as the con-clusive statement to the Agostini research as discussed i n this chapter: In the preceding sections we have shown that although .K i s not a universal constant, the em-p i r i c a l formula proposed by Agostini has a sound a n a l y t i c a l base and can be used to simplify the problem of estimating reach for advertising cam-paigns . In addition we have shown the nature and extent of the biases which r e s u l t from using the Agostini approach, and that simple procedures such as least squares regression can be used to obtain an estimate of K which i s appropriate for a given body of data. In summary, i t seems the only thing magical about K are: 1. It varies within a limited range. 2. It i s r e l a t i v e l y i n s e n s i t i v e to s p e c i f i c m edia—especially i f there are many issues and/or media i n the combination. 3. Its variance decreases sharply as D/A and the number of media i n the combina-tio n increase. 4. Estimates of reach s t a t i s t i c s are quite i n s e n s i t i v e to s p e c i f i c a t i o n errors i n K. Hence, i f K i s properly estimated for a given body of data, errors created by treating i t as a constant can usually be i g n o r e d . ^ The above l i t e r a t u r e can be related to i n t r a - p u b l i c a t i o n audience measurement i n three d i s t i n c t but interdependent ways. F i r s t , i t must be established that the same measures of dup-l i c a t e d audience and unduplicated audience are equally a p p l i -cable to i n t r a - p u b l i c a t i o n data as i n t e r - p u b l i c a t i o n data. This was discussed previously and i s fundamental to the r e l a -tionship between the l i t e r a t u r e of t h i s chapter and the objec-tives of the thesis. Second, i t must be shown that there i s a simple r e l a t i o n s h i p between net unduplicated audience and some e a s i l y calculable measure of audience duplication such as two-by-two audience combinations. This, i n e f f e c t , means finding some relationship between newspaper sections which i s the same as Agostini's " e q u i l a t e r a l hyperbolic curve." The above review of the "Agostini t r a d i t i o n " i s i n -tended to give at least a strong d i s p o s i t i o n toward the' existence of such a rel a t i o n s h i p . Third, i t must be demon-strated that the calc u l a t i o n of K, as described above, i s equally applicable to int r a - p u b l i c a t i o n newspaper data. The a r t i c l e of Marcel Marc c a l l i n g for re v i s i o n of K upwards for homogeneous combinations of magazines possibly has relevance for newspaper readership. Also Metheringham and Hofmans' concentration on duplication across several issues has prob-able implications for newspapers where r e p e t i t i v e and cumu-l a t i v e advertising effects are important. In short, once the relevance of Agostini's technique to in t r a - p u b l i c a t i o n data i s established, the establishment of 'special K' for newspaper audiences and audience segments should provide s i m i l a r problems as those discovered above. As noted before, this thesis w i l l attempt to establish the f i r s t of the above relationships, i n e f f e c t , a l l but proving the a p p l i c a b i l i t y of the Agostini research. The general objective of this thesis i s to show the superiority of segmented audience measures over aggregate audience measures. To the extent.that i t i s possible to demonstrate such a measure of unduplicated audience yet not provide a simple and workable method of c a l c u l a t i o n other than the rather laborious method of Chapter V or such studies as CESP, the proof of the hypo-theses would have l i t t l e p r a c t i c a l s i g n i f i c a n c e . Hence the above l i t e r a t u r e i s an important j u s t i f i c a t i o n of the thesis' obj ectives. CHAPTER IV HYPOTHESES AND THEORETICAL MODEL As s t a t e d i n the purpose of t h i s s t u d y , g e n e r a l aggregate a u d i e n c e f i g u r e s can be r e f i n e d by the use o f segmented audience measurement. The weakness of aggregate a u d i e n c e f i g u r e s i s t h a t t h e y a r e o n l y r e l i a b l e i f each r e a d e r reads the newspaper e x h a u s t i v e l y from c o v e r t o c o v e r . E x h a u s t i v e r e a d e r s h i p , i t can be argued, i s not a r e a l i s t i c d e s c r i p t i o n of-most i n d i v i d u a l s ' newspaper r e a d i n g h a b i t s i f f o r no o t h e r r e a s o n than t h e r e a d e r ' s l a c k of t i m e . T h e r e f o r e * t h e need f o r segmented audi e n c e measures. F o r t h e a d v e r t i s e r who w i s h e s t o p l a c e a campaign i n a number o f magazines, i t i s n e c e s s a r y t o r e c o g n i z e t h a t t h e r e may be m u l t i p l e magazine r e a d e r s h i p among t h e t o t a l a u d i e n c e he i s t r y i n g t o r e a c h and hence a s o u r c e o f d u p l i -c a t i o n o f exposure t o the campaign. U s i n g t h i s , t h e adver-t i s e r may w i s h t o maximize h i s r e a c h by s p r e a d i n g h i s cam-p a i g n t h r o u g h w i d e l y d i f f e r e n t i a t e d magazines o r he may w i s h t o maximize f r e q u e n c y f o r a s m a l l e r audience segment by * As an i n t e r e s t i n g e x t e n s i o n i t i s w o r t h n o t i n g t h a t t h e r e are o n l y t h r e e p o s s i b l e ways t o r e a d a newspaper: ( 1 ) e x h a u s t i v e l y — a method d i s c o u n t e d above, ( 2 ) randomly, o r ( 3 ) s e l e c t i v e l y . W h i l e f o r t h i s t h e s i s e i t h e r random o r s e l e c t i v e r e a d e r s h i p would show the d e s i r e d n u m e r i c a l r e s u l t s , o n l y s e l e c t i v e r e a d e r s h i p would a t t a c h any meaning t o t h o s e r e s u l t s . F o r t u n a t e l y , random b e h a v i o u r i s n o t c o n s i d e r e d t o be a human c h a r a c t e r i s t i c and f u r t h e r , the t h e o r y o f s e l e c -t i v e a p p e r c e p t i o n p r o v i d e s an i n t u i t i v e b i a s i n f a v o u r o f s e l e c t i v e r e a d e r s h i p . c o n c e n t r a t i n g and r e p e a t i n g h i s campaign i n o n l y one o r two magazines. He cannot maximize b o t h . T h i s e s s e n t i a l l y i s the l o g i c b e h i n d t h e r e s e a r c h done i n the A g o s t i n i t r a d i t i o n . The hypotheses s t a t e d below w i l l attempt t o expand t h i s r e s e a r c h by a p p l y i n g the same t r a d e - o f f concept t o the s e c -t i o n s w i t h i n an i n d i v i d u a l p u b l i c a t i o n . Hypotheses The hypotheses f o r t h i s t h e s i s are b e s t s t a t e d i n terms of an a d v e r t i s i n g campaign t o be p l a c e d i n a newspaper: (1) f o r a g i v e n a d v e r t i s i n g campaign, (a) r e a c h and f r e q u e n c y o f exposure f i g u r e s based on aggregate a u d i e n c e d a t a w i l l n o t r e p r e s e n t t h e t r u e measures of r e a c h and f r e q u e n c y f o r t h e campaign. DA ? DA agg r c UDA £ UDA agg c F 7^ F agg c DA = d u p l i c a t e d a u d i e n c e UDA = u n d u p l i c a t e d a u d i e n c e F = f r e q u e n c y o f exposure. (b) segmented a u d i e n c e measures of r e a c h and f r e q u e n c y w i l l be a c c u r a t e i f and o n l y i f the campaign does not cut across segmenta-ti o n boundaries. DA = DA where c e x secx c UDA = UDA secx c F = F secx c ( 2 ) the reach and frequency of exposure for any given campaign cannot be simultaneously maximized. f(F , UDA ) = 0 c c max max a S F = f ( UDA > Description of the Relevant Variables Sections: An advertiser, when placing an advertising campaign i n p r i n t media, can distinguish several d i f f e r e n t magazines or newspapers. S i m i l a r l y , i f he so wished, he could distinguish d i f f e r e n t sections within a magazine or newspaper. This study defines three types of sections. F i r s t , there are physical sections, the number of physical units into which the paper can be divided. Each of these units i s begun with a new front page, that i s , the f i r s t page of the second section i s l a b e l -led i n the paper as the 'second front page.' Next, there are indexed sections which are l i s t e d i n the index at the front of the paper such as sports and finance. There are often two 28. or three indexed sections i n each physical section. F i n a l l y , there are content sections which, although they can occur anywhere i n the paper are i d e n t i f i e d according to c a r e f u l l y defined categories of subject material. The model below and the research procedure c a l l for defining one of the above as sections for the purpose of th i s study. I t was decided to select indexed sections. Con-sider for a moment the role that a 'section' plays i n r e l a t i o n to the study's objectives and procedure. F i r s t , as w i l l be discussed shortly, i t serves as a mechanism for i d e n t i f y i n g audience segments. I t i s the unit by which the reader selects what he s h a l l read. Second, i t also serves as the 'vehicle' for the placement of an advertising campaign. Indexed sections, i t can be argued, represent the appropriate balance between these r o l e s . Physical sections, though e a s i l y distinguishable as vehicles, may not represent very d i s t i n c t audience segments because of the i r variety of content. Content sections, though representing the most appropriate unit for reading s e l e c t i v i t y and audience segmentation, lack easy physical i d e n t i f i c a t i o n for campaign placement. Indexed -sections, on the other hand, are e a s i l y i d e n t i f i e d p h y s i c a l l y and t h e i r content i s s u f f i -c i e n t l y r e s t r i c t e d to distinguish among them while i n them-selves remaining highly homogeneous. Indexed sections,_then i n t u i t i v e l y seem the most appropriate although the analysis could be performed using any of the three d e f i n i t i o n s . Readers: The advertiser i s of course concerned with the number of people who read the i n d i v i d u a l sections of the paper and as a r e s u l t have the opportunity to be exposed to his cam-paign. The readership associated with a section i s an audi-ence segment. Hence a reader must be defined. A decision on the degree of readership within a section has to be made i n order to designate a respondent as a reader or non-reader. For instance, i f an i n d i v i d u a l reports having read 50 percent or more of a section, he could be designated as a reader. S i m i l a r l y , the l e v e l could be set at 10 percent or 90 percent. There i s no'a.priori way to choose what th i s l e v e l should be independent of i n an investigation of the data. If 90 percent were designated, the number of q u a l i f y i n g respondents might be too small to form a r e l i a b l e sample. If 10 percent were chosen, the number of q u a l i f y i n g respondents might be too large to e f f e c t i v e l y distinguish one section readership from another. An appropriate balance i s needed and hence the desig-nation of readership l e v e l must await an investigation of the * results i n Chapter VI. Duplicated audience: The duplicate audience figure measures the aggregate of a l l readers for a combination of sections * Because the data used i n this thesis were previously co l l e c t e d i n another study, the analysis i s constrained by a pre-determined sample s i z e . Hence i t i s necessary to con-sider what e f f e c t varying readership l e v e l w i l l have on the data i n order to select one which w i l l provide s i g n i f i c a n t r e s u l t s . 30. across a number of weeks. To the advertiser i t i s the t o t a l number of exposures his advertisements receive. Each time a reader reads a section he contributes to the duplicate audience figure, no matter i f he has read other sections during the same week or the same section the week before. Unduplicated audience: The unduplicated audience or net reach figure measures the aggregate of a l l readers for a combination of sections or the whole paper across a number of weeks who have looked at the section or paper one or more times but are only counted once. To the advertiser t h i s i s the t o t a l number of people who have been exposed to his campaign at least once. Frequency of exposure: This i s the average number of times a reader i s exposed to a combination of sections or the whole paper. This figure can be found by d i v i d i n g the t o t a l number of exposures by the number of readers. In the above d e f i n i t i o n s what determines the sections or combinations of sections for which the measures are c a l -culated i s the combination of sections into which the adver-t i s e r places his messages. Hence when thi s paper refers to a combination of sections for the purpose of determining duplicate audience, unduplicated audience and frequency i t i s i n e f f e c t r e f e r r i n g to the measurement of exposure for a * mix of certain advertising locations within the newspaper. * Note that measures calculated for entire paper repre-sent one of the possible placement combination. 31. The Model The three variables described above, duplicated audi-ence, unduplicated audience and frequency of exposure are the fundamental tools of media audience measurement. The proof of the research hypotheses depends primarily on the derivation of these figures. The purpose of developing the following model i s to delineate the t h e o r e t i c a l relationships between the variables. For the purposes of the model d e f i n i t i o n s of sections and readership w i l l be assumed. In ef f e c t / then, there w i l l e x i s t a number of sections and a set of readers (audience seg-ment) associated with each. Also, for reasons of s i m p l i c i t y , the procedure w i l l be demonstrated for only one week or issue, a r e s t r i c t i o n which l a t e r must be relaxed. Suppose an advertiser decides to place i d e n t i c a l adver-tisements i n each section of a three section newspaper. Sup-pose further, that the readership pattern of that newspaper can be characterized as follows.(Fig. 3): Figure 3 S^ = set of readers associated with section 1. Certain readers read section 1 and nothing else, some only section 2, some only section 3, but some read two of the sections and others a l l three. To calculate the duplicate audience for the entire paper, the readerships of the sections are merely.aggregated. Hence the t o t a l number of exposures to the advertising campaign i s derived. However this figure obviously overestimates the net reach of the newspaper as some people who read more than one section are double or t r i p l e counted. Therefore the c a l c u l a t i o n of the net unduplicated audience must delete this source of error. Let DA = duplicated audience UDA = unduplicated audience S^, S 2, S 3 = readership of sections 1, 2 and 3, then, ( 1 ) D AS l , S 2 , S 3 = S l + S2 + S3 ( 2 ) U D AS l , S 0 , S , = S l + S2 + S3 " ( S 1 ° S2> 'l'"2' 3 - (s 2 n s 3) - (s 1 n s 3) + (s, n s, n s.) Frequency of exposure, as stated, i s the average number of times readers are exposed to the advertisement which i n th i s model was placed i n each section of the paper. 33. (3) F b l ' & 2 ' 3 The purpose of the above model i s only to demonstrate the concept of audience measurement on a section basis. The actual computation i s complicated by the fact that duplication over several weeks must be netted out of the duplicate reach figure before the above model can be applied, that there may be more than three sections and that the advertiser has the a b i l i t y to place his advertisements i n a widely varying pat-tern. Consider the cumulation of readers over the weeks. Each time a reader reports reading a section i n which an adver-tisement i s placed, he adds to the duplicate audience and frequency figures but not to the net unduplicated audience. The t h e o r e t i c a l procedure for subtracting t h i s source of dup-l i c a t i o n i s i d e n t i c a l to the model described above except that S^, S2 and would represent the readership of the same sec-ti o n over three successive weeks. Some people only read the section once over the three weeks, some twice, some a l l three * times. * In the actual method of analysis i n Ch. VI t h i s pro-cedure i s carr i e d out before c a l c u l a t i n g i n t e r - s e c t i o n d u p l i -c a t i o n — t h u s the o r i g i n a l model pp. 31-33 s t i l l applies. (tot a l number of exposures)S^,S 2,S^ (number of readers) S^ rS-^ D A S S S b l ' 2' 3 UDAQ 1' 2' 3 34. The fact that the actual analysis may involve more than three sections necessitates generalization of the model and l a t e r the method of analysis to n sections. The generali-zed model follows the same format as above: ( 1 ) D AS l , S 9 , S , S = S x 1' 2' 3 ' n x=1 (2) UDA 1' 2' 3''• • ' n n 2 S - i x x=l n-1 n - [ 2 2 ( s n S ) ] x=l y=x+l Y n-2 n-1 n + [ 2 i 2 (s x n s n sz) ] x=l y=x+l z=y+l y n n- (n-1) - (-D [ 2 x=l n-(n-2) n-(n-3) 2 2 y=x+l z=y+l n 2 (sv n s A s ... nsj ] k=n x y z k DA_ c c q (3) F_ _ . _ _ b l f b 2 / b 3 > . . . , b n b 1,b 2,b 3,...,b n y Referring again to the three section model, suppose now that there i s a second time period. Assume further that an advertiser places two advertisements i n the f i r s t two sections for week 1 and an advertisement i n each of the f i r s t and t h i r d sections for week 2. F i r s t , c a l c u l a t e the duplicate audience of the campaign i n week 1: (5) DA = DA , . , + DA ^ . , 1 secl,wkl sec2,wkl where c e 1+2 i n week 1 Next, calculate the duplicate audience of the campaign i n week 2: (6) D AC2 D Asecl,wk2 + DAsec,3,wk2 where c e 1+3 i n week 2 The duplicate audience of the campaign i s (7) DA = DA + DA c c x c 2 D A s e c l , w k l + ^ s e c l ,wk2 + D Asec2,wkl + ^sec3,wk2 By contrast, calculate the aggregate duplicate audience of the paper, regardless of the campaign placement: (8) DA = DA , . , + DA 0 . , + DA agg secl,wkl sec2,wkl sec3,wkl + D Asecl,wk2 + D Asec2,wk2 + D Asec3,wk2 3 6 . Although present audience measurement techniques recognize the concept of duplication, the procedures tend to aggregate and average data over sections and over time thus, i n a sense, hiding what could be s i g n i f i c a n t data f o r better campaign de-sign. Hence hypothesis 1(A) tests whether or not a less than exhaustive campaign placement (Eq.7) w i l l provide the same reach and frequency as would an exhaustive placement over that combi-nation of sections and weeks designated i n the campaign (Eq.8). As the objective of t h i s thesis i s to demonstrate that segmented audience data provide a superior basis for campaign measurement than aggregate audience data, the inference i s c l e a r l y . t h a t the exposure results to the above campaign should be calculated on the basis of the audience segments associated with each section. This method of c a l c u l a t i o n would obviously be superior i f the campaign consisted of two advertisements placed successively i n section 1; however, i t consists of four advertisements and cuts across three sections. Clearly then, the exposure results for section 1 do not represent the actual exposure to the campaign. Furthermore, the exposure results for section 2 do not r e f l e c t the exposure of campaign i n that sections because only one of the available weeks i s used. Hence hypothesis 1(B) r e f l e c t s the l i m i t a t i o n s of the segmented audi-ence concept. However, the proof of hypothesis 1(B) does not negate the use of audience segments as w i l l be demonstrated i n the findings of this study. * Equations 7 and 8 refer to only duplicated audience--the results w i l l be demonstrated for unduplicated audience and frequency as well. As a co r o l l a r y to the above analysis, i t i s obvious that i f the concept of segmented audiences describes reader-ship patterns c o r r e c t l y , the advertiser has a decision to make concerning the net reach of the campaign and the f r e -quency of exposure of the audience to i t . He can maximize unduplicated audience by spreading his campaign through the di f f e r e n t sections. A l t e r n a t i v e l y , he can maximize frequency by concentrating the campaign i n a single section. Hence * hypothesis 2. * The appropriate trade-off between reach and frequency i s a function of the objectives of an advertising campaign. For example, suppose an advertiser designs a campaign with the objective of reaching a lim i t e d audience segment. His strategy may be to stimulate learning by repeated exposure of his advertising messages to that audience. In placing the campaign he would be concerned with selecting that section of the newspaper most heavily read by the relevant audience seg-ment. He could then concentrate his campaign i n that section and thereby maximize the average frequency of target audience exposure. A l t e r n a t i v e l y , an advertiser may design a campaign which does not d i f f e r e n t i a t e among p a r t i c u l a r audience seg-ments. The strategy would be to expose as broad a percentage of t h e t t o t a l newspaper audience as possible to the campaign. By placing advertisements across many sections, e s p e c i a l l y those with low overlapping readership, he would be attempting to maximize reach while s a c r i f i c i n g duplication of exposure to the campaign. In these or sim i l a r s i t u a t i o n s , the net reach and frequency of exposure of alternative campaign placements w i l l obviously have significance for the design of an o v e r a l l adver-t i s i n g strategy. I t i s the purpose of hypothesis 2 to show that a trade-off between reach and frequency does exi s t and does bear upon the advertisement placement decision process. However i t i s beyond the scope of this paper to li n k placement strategies to s p e c i f i c advertising objectives other than to demonstrate that such a l i n k e x i s t s . CHAPTER V METHOD OF ANALYSIS This chapter w i l l describe i n p r a c t i c a l terms the procedure for drawing the necessary information from the data base and a method of analysis consistent with the t h e o r e t i c a l model outlined i n Chapter IV. There were 1,220 respondents to the questionnaire and 35 computer source cards associated with each respondent. Cards 25 to 35 contain demographic and psychological data which are not of d i r e c t relevance to this study. Cards 1 to 24 are arranged i n six groups of four, each group representing answers to questionnaires on the broadcast, as well as the p r i n t media, for a s p e c i f i c week. Only the l a s t two cards of each group contain newspaper data, for instance cards 3 and 4 of group (week) 1, 7 and 8 of group (week) 2, etc. In raw form the data were arranged by type of card rather than by person. As i t i s necessary for this study to aggregate the t o t a l readership by section, the cards f i r s t had to be sorted. The r e s u l t i n g arrangement grouped the data for respondent 1 together and i n order, followed by the data for respondent 2 and so forth. I t was noted before that the respondents were to indicate whether or not they read or saw the paper, i n d i v i d u a l pages and i n d i v i d u a l quarter pages. Their responses: yes, no or no response, are indicated on the appropriate data card. I t was also noted before that this study w i l l desig-nate indexed sections as the appropriate d e f i n i t i o n of sec-tio n . Each such section i s represented on the data cards by a group of quarter page responses. By aggregating the posit i v e questionnaire responses within a 'section' and comparing the percentage to the designated readership-nonreadership l e v e l , i t can be determined whether or not a respondent can be c l a s s i f i e d as a reader: C . = 0 i f Q q < R. . 1 ] Q- 1 3 C. . = 1 i f Q q > R. . ID Q- " x D where Q = t o t a l quarter page responses for a section q = p o s i t i v e responses R. . = readership l e v e l . iD * The data tabulation procedure can now be c l a r i f i e d by the use of a matrix. The c e l l s i n the matrix (C..) are f i l l e d iD with a 1 or 0 depending respectively on whether or not the reader i s above the required' readership l e v e l . Note that the matrix (Fig. 4) compares sections and people for a single time period. Later the results w i l l be aggregated over the weeks. 40. Figure 4 Individual Section Readership sections respondents^Sj^^ 1 2 N-l N person 1 C. . C C. . C. . ID 13 ID ID person 2 C. . " " ID person T - l " " " person T "- ^ . . . . . " S l S 2 S N . - 1 SN By t o t a l l i n g the columns the number of respondents who read section 1 can be determined (S-^ ) , the number of respon-dents who read section 2 ( S 2 ) , and so forth. Referring back to the t h e o r e t i c a l model i n Chapter IV i t w i l l be remembered that the research method c a l l e d for deleting duplication caused by multiple section readership. Hence, matrices similar to the above can be developed for multiple section readers: 41. Figure 5 Multiple Section Readership sections respondentsS. 1&2 1&3 l&N 2&3 N-1,N person 1 C. . 1D C. . . . ID . C. . ID, C. . . . ID . C. . ID person 2 C. . ID II . 11 II II person T - l II . ii II II II person T II ii II 11 II S1.2 S l , 3 * ' • S1,N S2,3' . * * SN-1,N By t o t a l l i n g the columns the following information can be derived: S, 0 respondents read both sections 1 and 2 S, respondents read both sections 1 and 3 X , j ^N-l N r e s P o n i ^ e n " t s r e a d both sections N-l and N Sim i l a r l y , matrices for 3,4,...,N section readership can be tabulated. F i r s t , t h e n , t h e d u p l i c a t e d a u d i e n c e must be c a l c u l a t e d o v e r w = 6 w e e k s . F o r a s i n g l e s e c t i o n (#1 i n F i g . 4) t h e a u d i e n c e was S^. The d u p l i c a t e a u d i e n c e f o r t h a t same s e c -t i o n o v e r "w" weeks i s : DA , = E S, s e e l 1 w S i m i l a r l y t h e d u p l i c a t e a u d i e n c e may be c a l c u l a t e d f o r any c o m b i n a t i o n o f s e c t i o n s o r a l l t h e s e c t i o n s i n t h e news-p a p e r o v e r "w" w e e k s : DA , „ s e e l , s e c 2 , . . . , s e c N £ S 1 + £ S 2 + . . . + £ s w w w N Now, t h e c a l c u l a t i o n o f t h e u n d u p l i c a t e d a u d i e n c e i s more c o m p l e x . I t i n v o l v e s c a l c u l a t i n g t h e n e t r e a c h o f a s e c t i o n o v e r t h e weeks and t h e n t h e n e t r e a c h o f a c o m b i n a t i o n o f s e c t i o n s . The n e t r e a c h o f a c o m b i n a t i o n o f s e c t i o n s w i l l n o t s i m p l y be e q u a l t o t h e sum o f t h e n e t r e a c h o f t h e s e c t i o n s due t o i n t e r - s e c t i o n d u p l i c a t i o n . T h u s , t h e f o l l o w i n g two s t e p p r o c e d u r e i s d e s i g n e d t o g e t r i d o f b o t h s o u r c e s o f d u p l i c a t i o n , t h a t w h i c h o c c u r s a c r o s s t h e weeks and t h a t w h i c h o c c u r s a c r o s s s e c t i o n s . The b l o c k o f d a t a now a p p e a r s as i n F i g u r e 6: Figure 6 Consider f i r s t , single sections over 'w' weeks. Each c e l l i n the block w i l l have been f i l l e d with either 1 or 0 i t w i l l be r e c a l l e d indicating readership or non-readership. (C. . = 1 or 0) I j X. . = 0 i f E C . = 0 J W J This procedure compiles the results over the *w' weeks and eliminates that source of duplication. Once completed, the r e s u l t i s simply a section/people matrix indicating whether or not each person read a section at least once. I n o r d e r t o c a l c u l a t e t h e t o t a l n e t r e a d e r s h i p o f t h e s e c t i o n s , i t i s o n l y n e c e s s a r y t o sum X.. o v e r T ( t h e number o f r e a d e r s ) N T E X. . 1=1 However as n o t e d a b o v e , m e r e l y a g g r e g a t i n g t h e n e t r e a c h o f t h e s e c t i o n s w i l l n o t p r o d u c e t h e u n d u p l i c a t e d r e a c h o f t h e p a p e r . R e s p o n d e n t s who r e a d m u l t i p l e s e c t i o n s a r e s t i l d o u b l e c o u n t e d . T h e r e f o r e i t i s n e c e s s a r y t o c a l c u l a t e m u l -t i p l e s e c t i o n r e a d e r s h i p b y g o i n g t h r o u g h a s i m i l a r p r o c e d u r e as b e f o r e : people, F i g u r e 7 X. iD S e c t i o n p a i t s 0 i f Z C . w 1 j = 0 w e e k s X, ID NR 1 i f Z C . 1 0 W l j T E X. p a i r s ± = 1 i ] The process i s similar for a l l combinations of sections. The computation of the j o i n t reach of a combination of sections yi e l d s the duplicate exposure and hence the i n t e r -section described i n the th e o r e t i c a l model. For any combina-t i o n of sections or for the whole paper i t i s then possible to compute the unduplicated audience. Assuming a three sec-ti o n newspaper: U D A1,2,3 = N R 1 + N R 2 + N R 3 " N R l + 2 ~ N R l + 3 - NR 2 + 3 + NR 1 + 2 + 3* The frequency of exposure i s then simply calculated from duplicated and unduplicated audience DA.. , , E S]_ + E S 2 + E S 3 F _ : -1,2,3 _ w w w UDA 1 i 2 f 3 NR 1 + NR 2 +NR 3-NR 1 + 2-NR 1 +3-NR 2 +3 +NR 1 + 2 +3 This figure can be calculated for i n d i v i d u a l sections as well as for any combination of sections. Hence the deriva-t i o n of the indices fundamental to the research hypotheses. It remains only to demonstrate the relationships set out i n the hypothesis. * Some confusion may re s u l t when comparing th i s equation with equation (2) of the model i n Chapter IV. In the model S was substituted for NR above. The model, i t w i l l be remembered, was concerned with only one time period. The method of analy-s i s derived S on page 34 but i t i s only equal to NR i n a single time period. NR d i f f e r s from S because i t has eliminated dup-l i c a t i o n over the weeks. Appendix . A l o g i c a l question concerning the above method of analy-s i s i s how does one i s o l a t e the exposure results of a campaign which uses only cer t a i n sections and issues of the newspaper. The method appears to allow only the results of large combina-tions of sections and weeks. The answer i s simply that the computer programming i s adjusted for each campaign so that only the exposure results of the relevant sections are read by the computer. The program proceeds to compile the exposure results as i t would do i f a l l the sections and issues were included. The e f f e c t given at the end i s that none of the respondents read any of the section/week c e l l s i n which there was no campaign placement. This enables the recording of the exact exposure results for the c e l l s - i n which the campaign was placed. CHAPTER VI SUMMARY OF RESULTS This chapter presents and discusses the exposure results to a series of hypothetically placed advertising campaigns. However, before r e l a t i n g the actual findings, two problems must be considered which were l e f t unresolved because of the need to examine the data before drawing con-clusions: the missing data and the readership l e v e l . The Problem of the Missing Data As outlined in Chapter I I , the data used i n this study were not co l l e c t e d with the exact objectives of this thesis i n mind. The vehicles of campaign placement as described previously are the sections and issues of the newspaper. Un-fortunately, the abridged questionnaire did not reproduce every quarter page i n every week. Frequently, i n fact, there was no representation of a section i n a p a r t i c u l a r week. Hence the questionnaire did not always provide an opportunity for the respondent to indicate whether or not he had been exposed to p a r t i c u l a r indexed sections. Consider again for a moment Fig. 6 i n the method of analysis. I t i s possible to draw from that block a section/week matrix describing the available placement opportunities for an advertising campaign. If there were, for example, no pages from the sports section i n week 3 48. reproduced i n the questionnaire, there would be a blank i n the section/week decision matrix. The problem i s further complicated by the fact that panels A, B and C received s l i g h t l y d i f f e r e n t newspapers and hence the missing section/ week c e l l s were not the same i n each panel. This does not seriously a l t e r the l o g i c of the research procedure. The aggregate exposure results for the newspaper could be redefined as the aggregate for the sections and issues which were available and campaign exposure would s t i l l be correct as long as the campaigns were placed i n the section/ week c e l l s for which results are available. However this constraint proved to be an awkward l i m i t a t i o n . When the three panels were merged the r e s u l t i n g decision matrix appeared as in F i g . 8. Where 1 S^W_.' appears results could be tabulated across a l l three panels for that c e l l . Where 'X' appears the questionnaire was lacking i n at least one of the three panels. Figure 8 Sections' Sports Finance Women Entertainment 1 S l V 1 l X X S 4 W l 2 S.^ X S3W2 X r 7 . 3 S,W0 X X X Weeks 1 3 (issues) 4 3 ^ X 5 S,WC S„W C S.,Wr X 1 5 2 5 3 5 6 S-,Wr X S-,W,. X 1 6 3 D The choice of the four sections: sports, finance, women's and entertainment appeared to provide the most com-plete results of the available indexed sections. The above matrix obviously l i m i t s seriously the a b i l i t y to develop hypothetical advertising campaigns which an advertiser might l o g i c a l l y place i n a newspaper. It was decided then to conduct the analysis on a seg-ment of the available data. While providing more complete exposure results this decision had to be made at some s a c r i -f i c e to sample si z e . Panel A was chosen and i s depicted i n Fi g . 9. Figure 9 Sections Sports Finance Women Entertainment 1 S 1 W 1 S 2 W 1 S 3 W 1 S 4 W 1 2 S1 W2 s2w2 S3 W2 S4 W2 Weeks 3 S1 W3 s2w3 S3W3 \X (Issues) 4 S1 W4 S2 W4 S3 W4 X 5 S 1 W 5 s2w5 s3w5 X 6 S1 W6 S 2 W 6 S 3 W 6 X The sections sports, finance and women's over the six weeks provide the largest complete block of data available. The entertainment section was included to demonstrate the 50. v a l i d i t y of the procedure even though there are gaps ex i s t i n g in the data. The fourth section also provides more variety i n the placement of advertising campaigns. Panel A provides 20 section/week alternatives for placement of advertisements as compared to 2 4 i f the data were complete. The comparison of campaign results w i l l remain legitimate as long as none of the campaigns make use of the entertainment sections i n weeks three to s i x . The matrix i n F i g . 8 for a l l three panels pro-vides only 13 placement alternatives for advertisements. As noted before there were 1,220 respondents to the questionnaire for panels A, B and C combined. For panel A the sample size i s 402. Readership Level The exposure re s u l t s of a l l the hypothetical campaigns used i n this thesis, as well as the aggregate exposure for panel A were calculated on the basis of four d i f f e r e n t reader-ship l e v e l s : .1, .25, .5, .75. The reasons for t h i s were discussed i n Chapter IV. The hypotheses were tested primarily upon the basis of 16 campaigns representing alternative place-ments of six advertisements. To select the most appropriate readership l e v e l the net reach results for the 16 campaigns were used. The table below (Table I) indicates the range within which the net reach of most of the campaigns f e l l as a percen-tage of sample s i z e . TABLE I Net Reach as Percent of Readership Level Panel A for six-ad Campaigns  .1 7 8 - 8 6 Percent .25 68 - 82 Percent .5 4 7 - 7 4 Percent .75 36 - 59 Percent At a readership l e v e l of .1, then, at least 78 percent and not more than 86 percent of panel A was exposed to each campaign. The d i f f i c u l t y with such a high l e v e l of exposure i s that because of the extensive duplication of readers across the newspaper, i t would not allow the i d e n t i f i c a t i o n of audi-ence segments that were s i g n i f i c a n t l y d i f f e r e n t from each other. From another point of view, i t would be d i f f i c u l t to j u s t i f y a respondent who reported having been exposed to 10 percent of a section as a reader of that section. The place-ment of an advertisement i n a section could hardly guarantee r e s u l t s . A readership l e v e l of .25 suffers the same kind of d i f f i c u l t i e s but of course to a lesser degree. At levels of .5 and .75, the i d e n t i f i c a t i o n of d i s t i n c t audience segments becomes more r e a l i s t i c . Also, at these levels the advertiser can be sure that those people who are exposed to the section are i n fact actual readers of that section. Another input-into the selection of the most appropriate readership l e v e l i s , of course, the quality of the advertisement. To the extent that i t i s a large, eye-catching advertisement a readership l e v e l of .25 or perhaps .1 may be enough as far as the advertiser i s concerned. As t h i s thesis depends upon hypothetical advertising campaigns, the factors discussed above have no i d e n t i f i a b l e influence. Hence, a l l things equal, i t was most appropriate to select .5 as the readership l e v e l . A l l r e s u l t s presented in the remainder of t h i s chapter are thus calculated at the .5 l e v e l . The results obtained at other readership levels are presented i n the appendices to the thesis. Presentation of Results Hypothesis 1(A) stated that reach and frequency of exposure figures based on aggregate audience w i l l not repre-sent the true measures of reach and frequency for the campaign. The c a l c u l a t i o n of measures based on aggregate data assumes that the campaign has achieved complete audience coverage. The proof of the hypothesis l i e s i n demonstrating that a less than exhaustive campaign placement w i l l not provide the same reach and frequency as would an exhaustive placement over that combination of sections and issues designated i n the campaign. Thus a campaign consisting of twelve advertisements placed in some l o g i c a l pattern across four sections and six weeks would not provide the same results as a campaign placed i n each section every week. Six hypothetical advertising campaigns of sizes s i x , twelve and eighteen advertisements were selected from the array of placement p o s s i b i l i t i e s i n panel A. They are des-cribed below and use the same notation as i n Fig. 9: Campaign 1(A): Six successive advertisements placed for six issues i n one section (sports) S 1 W 1 ' S 1 W 2 ' S 1 W 3 ' S l V S l V S1 W6 Campaign 1(B): - Two advertisements placed i n each of three sections, one i n the f i r s t issue and one i n the sixth issue S 1 W 1 ' S1 W6 S 2 W 1 ' S2 W6 S 3 W 1 ' S3 W6 Campaign 2(A): Four advertisements placed i n each of three sections, two i n the f i r s t two issues and two i n the la s t two issues S 1 W 1 ' S 1 W 2 ' S 1 W 5 ' S1 W6 s2w2, s2w2, s2w5, s2w6 s 3w l f s3w2, s3w5, s3w6 Campaign 2(B): Two advertisements placed i n each of three sections i n the f i r s t two issues and six advertisements placed i n the remaining section for six successive issues. 54. S 1 W 1 ' S1 W2 S 2 W 1 ' S2 W2 s 3w 1 ,/;s 3w 2 , s 3 w 3 / s 3w 4 , s 3 w 5 / s3w6 s 4 w 1 # s4w2 Campaign 3(A): Six advertisements placed i n each of three sections for six successive issues s l W ] L , S l w 2 , S l w 3 , S l w 4 , . S l w 5 , s^6 S 2 W 1 ' S 2 W 2 ' S 2 W 3 ' S2 W4; S 2 W 5 ' S2 W6 s ^ , s 3w 2 , s 3w 3 , s 3w 4, s 3w 5, s3w6 Campaign 3(B): Six successive advertisements placed for s i x successive issues i n one section, f i v e advertisements placed i n each of two sedtions excepting issue 3, two adver-tisements i n the fourth section for the f i r s t two issues. S l w i f S l w 2 , S ; L w 3 , S ] W 4 , S l w 5 , S l w 6 S 2 W 1 ' S 2 W 2 " S 2 W 4 ' S 3 W 5 ' S3 W6 s 3 W l , s 3w 2 , s 3w 4 , s 3w 5 , s3w6 S 4 W 1 ' S4 W2 Aggregate: Exhaustive placement of advertisements i n each available section and issue of the newspaper. s^, s^, s1w4, s1w5, S-jWg s2wr s2w2, s2w3, s2w4, s2w5, s2w6 S 3 W 1 ' S 3 W 2 ' S 3 W 3 ' S 3 W4' S 3 W 5 ' S3 W6 s4wIf s4w2 Comparison of the exposure results (Table II) demon-strates conclusively that the use of aggregate data over-estimates exposure results for actual advertising campaigns. The aggregate based measure of duplicated audience i s almost four times as great as actual duplicated audience for campaigns of s ix advertisements.[1(A) and 1(B)]. The aggregate measure of unduplicated audience exceeds those for campaigns 1(A) and 1(B) by at least 30 percent and a simi l a r camparison for f r e -quency y i e l d s a discrepancy of at least 125 percent. As the campaign size increases to twelve and then eighteen advertise-ments the error caused by using aggregate data decreases accor-dingly. This i s to be expected as the campaign size approaches an exhaustive placement of advertisements i n each available section and issue. However the degree of error r e s u l t i n g from the adjustment of the campaign size i s of secondary importance. The fundamental conclusion to be drawn from Table II i s that aggregate audience based measures of duplicated reach, undup-l i c a t e d reach and frequency of exposure c l e a r l y overestimate the exposure to actual advertising campaigns. Hence according to the c r i t e r i a outlined on page 52, hypothesis 1(A) has been proven. However consider for a moment a campaign which con-s i s t s of twenty advertisements and i s placed exhaustively through the available section/issue combination. TABLE II Campaigns 1 (A) 1(B) 2(A) 2(B) 3(A) 3(B) Aggregate Duplicated Audience 508 596 1121 1398 1680 1831 2018 Unduplicated Audience 193 256 290 320 306 330 331 Frequency of Exposure 2.64 2.33 3.87 4.37 5.49 5.55 6.10 In what manner would measures based on aggregate data not r e f l e c t the true measures for t h i s s p e c i f i c campaign? The answer l i e s not i n what the aggregate measures state but i n what they do not state. In fact they are accurate but they imply equitable exposure to the d i f f e r e n t sections of the newspaper which i s not the case. Table III gives a breakdown of the newspaper exposure 'into the audience segments associated with each section. At thi s point a l l that need be said i s that the aggregate based figures are misleading i n the sense that they average out the rather considerable discrepancies among the audience segments. Some of the implications of Table III w i l l be discussed at a l a t e r point. Extending this analysis to the results given i n Table I I , there i s further support to hypothesis 1(A). Audience measurement based on aggregate data, then, do not represent the true measures of an adver-t i s i n g campaign. I t has been implied that measures based on segmented audience data give better results but c l e a r l y there i s a l i m i t a t i o n involved i n such a conclusion. This i s the subject of the next hypothesis. TABLE III * Sports Finance Women's Entertmt. Aggregate Duplicated Audience 508 439 733 338 2018 Unduplicated Audience 193 170 213 222 331 Frequency of Exposure 2.64 2.58 3.44 1.52 6.10 Hypothesis 1(B) stated that segmented audience measures of reach and frequency w i l l be accurate i f the campaign does not cut across segmentation boundaries. I t i s necessary then to demonstrate the exposure results of campaigns which are placed exclusively for p a r t i c u l a r audience segments as well as campaigns placed to reach several audience segments. The following set of hypothetical campaigns i s designed to provide comparable r e s u l t s . Again the same notation as i n F i g . 9 i s used to describe the campaigns: * I t w i l l be remembered that results for the entertain-ment section can only be based on two issues. Campaign 1: S^W^ S . ^ , S-JW-J, S^^, S . ^ , S-jWg Campaign 2: s 2 w i ' S 2 W 2 ' S 2 W 3 ' S2 W4' S 2 W 5 ' S2 W6 Campaign 3: S 3 W i ' S3 W2' S 3 W 3 ' S3 W4' S 3 W 5 ' S3 W6 Campgign 4: S^W^ S ^ V i ^ , S2W-L, S2W2, S2W3 Campaign 5: s i w i ' S 1 W 2 ' S 2 W 1 ' S 3 W 1 ' S 3 W 2 ' S 4 W 1 Note that each campaign consists of s i x advertisements and that numbers 1, 2 and 3 are placed exclusively i n a par-t i c u l a r section and hence appeal to the respective i n d i v i d u a l audience segments while numbers 4 and 5 are placed across sections and appeal to more than one audience segment. To prove that measures based on segmented audience data are not accurate for campaigns placed across audience segments, i t i s only necessary to compare exposure results for a l l f i v e campaigns, as i n Table IV. TABLE IV Campaigns 1 2 3^  4 5 Duplicated Audience 508 439 733 473 662 Unduplicated Audience 193 170 213 188 267 Frequency of Exposure 2.64 2.59 3.44 2.52 2.48 The fact that audience segment based exposure results provide accurate estimates for campaigns 1, 2 and 3 i s of course a tautology as they are the same thing (compare Tables * III and IV). Segmented audience based data would however be useful in obtaining results for campaigns of less than six advertisements placed i n a single section once adjusted for the number of placements. To the extent that the exposure results for campaigns 1, 2 and 3 do not provide any i d e n t i f i -able estimate for the results of campaigns 4 and 5, the fact that segmented audience data w i l l not provide accurate results for campaigns placed across segments i s proven. This i s the signi f i c a n c e of hypothesis 1(B). Application of the segmented audience concept obviously then has no meaning i n terms of the t o t a l exposure to campaigns 4 and 5. Hence hypothesis 1(B) i s proven. However i t i s s t i l l a highly relevant concept and i t s application i s the genesis of the proof of hypotheses 1(A) and 1(B). Consider a compari-son of campaigns 1 through 5 over the four audience segments that have been i d e n t i f i e d . Some rather considerable' v a r i a t i o n can be i d e n t i f i e d : Campaigns 1, 2 and 3 are placed exclusively i n the sports, finance and women's sections respectively--hence the exposure to those campaigns (Table IV) must be the same as the exposure results of the respective sections (Table III) . 60. TABLE V Sports 1 2 3 4 5 DA 508 - - 284 2 05 UDA 193 - - 164 151 F 2.64 - - 1.75 1.36 Finance 1 2 3 4 5 DA - 439 - 189 64 UDA - 170 - 107 64 F - 2.58 - 1.77 1.00 Women 1 2 3 4 5 DA - - 733 - 229 UDA - - 213 - 152 F - - 3.44 - 1.51 Entertainment 1 2 3 4 5 DA - - - - 16 4 UDA - 164 F - - -•: - 1.00 The above table depicts the exact e f f e c t each campaign would have i n each audience segment. Despite the obvious differences among then, measures based on aggregate data, as can be seen i n Table I I I , would have indicated the same results for each campaign. To summarize, the advantage of exposure measurement on a segmented audience basis i s that while providing accurate o v e r a l l campaign measurement, the method also allows i d e n t i -f i c a t i o n of the net reach and frequency of exposure for each section into which the campaign i s placed. As a by-product the method also generates the duplication of exposure among audience segments. I t i s perhaps useful to trace this process through for campaign 4 before going on to hypothesis 2. TABLE VI Campaign 4 Duplicated audience DA (sports) = 284 DA (finance) = 189 DA (campaign) = 4 73 Net reach UDA (sports) = 164 UDA (finance) = 107 duplication of readers between sports and finance = 83 UDA (campaign) = (164 + 107)- 83 - 188 Frequency of exposure F (sports) = -T-|4 = 1.75 189 F (finance) = -. -,_ = 1.77 473 F (campaign) = 1 Q Q - 2.52 16,4 107 188 Hypothesis (2) stated that reach and frequency of exposure for any given campaign cannot be simultaneously maximized. To test this beyond question i t would be neces-sary to develop the complete set of placement alternatives for an advertising campaign of a given size. However con-sidering a six advertisement campaign and the twenty possible placement positions given i n F i g . 9 the number of combinations becomes unmanageable: 20! 6! (20 - 6) ! = 38,760 As a substitute, i t i s r e a l i s t i c to develop a set of campaign placements which r e f l e c t s extremes of concentration and dispersion of advertisements while remaining l o g i c a l alter-natives for an advertiser. They are presented below, again using the notation of Fig . 9. 1. 2. 3. 4. 5. 6. 7. 8. S 1 W 1 S 2 W 1 S 3 W 1 S 1 W 1 S 1 W 1 S 1 W 5 S 1 W 1 S 2 W 1 S1 W2 s2w2 s3w2 S1 W2 S1 W2 S1 W6 S1 W4 s2w2 S 1 W 3 s2w3 s3w3 S 1 W 3 S 2 W 1 S 2 W 5 S 2 W 1 S 3 W 1 S1 W4 S2 W4 S3 W4 S 2 W 1 s2w2 S2 W6 S2 W4 s3w2 S 1 W 5 s2w5 s3w5 s2w2 s3w1 s 3w 5 S 3 W 1 S 4 W 1 S1 W6 S2 W6 S3 W6 s2w3 s3w2 S3 W6 S3 W4 S4 W2 63. Random Random 9. 10. 11. 12. *13. *14. 15. 16. S 2 W 1 S2 W4 S 1 W 1 S 1 W 1 S 1 W 3 S1 W2 S 1 W 1 S 1 W 1 s2w2 s2w5 S1 W2 S 2 W 1 S1 W5 S 1 W 5 S 2 W 1 S1 W2 s2w3 S 2 W 6 S 2 W 1 s2w2 S2 W4 S 2 W 1 S 3 W 1 S 1 W 3 S 3 W 1 S3 W4 S 3 W 1 S 3 W 1 S 3 W 1 s2w3 s3w2 s2w2 s3w2 s3w5 s3w2 S 4 W 1 s3w3 S2 W4 s3w3 s3w2 s3w3 S3 W6 S 4 W 1 S4 W2 S3 W6 S 4 W 1 S 4 W 1 S4 W2 If the above campaigns are examined i t w i l l be noted that some concentrate heavily i n one section, some i n certai n combinations of sections. Others have been placed so as to be widely dispersed as possible, some have been placed with no regard to concentration of advertisements i n audience seg-ments, two campaigns were selected randomly. The results for each of these campaigns are given i n Table VII. Note that the results indicate that certain campaigns are superior to others both on the basis of net reach and frequency. However where frequency i s maximized at campaign 3, reach i s not. Any attempt to increase the reach of th i s campaign without inc-reasing the number of advertisements cannot be accomplished except at expense to the frequency of exposure. S i m i l a r l y * campaign 12 maximizes net reach but not frequency. See Table VIII: campaigns of Table VII re-arranged according to increasing net reach. 64. TABLE VII Campaigns Duplicated Audience Net Reach Frequency of Exposure 1. 508 193 2.64 2. 439 170 2.59 3. 733 213 3.44 ** 4. 473 188 2.52 5. 556 248 2.24 6 . 565 240 2.35 7. 596 256 2.33 8. 689 269 2.56 9. 538 216 2.49 10. 634 238 2.66 11. 662 267 2.48 12. 699 281 * 2.49 13. 624 241 2. 59 14. 540 239 2.26 15. 713 276 2. 58 16. 642 274 2.34 Hypothesis (2) i s a c o r o l l a r y to hypothesis (1) and hence i s true by d e f i n i t i o n . However i t i s important i n that i t distinguishes another d i f f i c u l t y i n the use of aggregate data. Aggregate data based measures do not allow the i d e n t i -f i c a t i o n of a trade-off between the reach and frequency of a set of advertising campaigns. The selection of the appropriate 65. trade-off i s of course a function of the advertiser's objec-t i v e s . He may not wish to maximize either reach or frequency but he should be aware that his selection of an advertising campaign involves some degree of s a c r i f i c e of one or both of these exposure variables. TABLE VIII Duplicated Frequency of Campaigns Audience Net Reach Exposure 2. 439 170 2.59 4. 473 188 2.52 1. 508 193 2.64 3. 733 213 3.44 ** 9. 538 216 2.49 10. 634 238 2.66 14. 540 239 2.26 6. 565 240 2.35 13. 624 241 2.59 5. 556 248 2.24 7. 596 256 2.33 11. 662 267 2.48 8. 689 269 2.56 16. 642 274 2.34 15. 713 276 2.58 12. 699 281* 2.49 * Re: footnote page 37. CHAPTER VII CONCLUSIONS The conclusions to thi s study, i n the simplest sense, coincide exactly with the objectives of Chapter I. The proof of hypothesis 1 demonstrates that measures of duplicated audience, unduplicated audience and frequency of exposure based on segmented audience data are superior to those same measures based on aggregate audience data. At the same time, the l i m i t a t i o n s of segmented audience measures are acknow-ledged for advertising campaigns which cut across audience segments. However, the results of Chapter VI demonstrate the a p p l i c a b i l i t y of the segmented audience concept even within the constraints of those l i m i t a t i o n s . Hypothesis 2 demon-strates that for a set of alternative placements of an adver-t i s i n g campaign, the reach and frequency measures cannot be simultaneously maximized. However i t i s not the intention of this chapter to merely summarize the l a s t one. Its objective i s to draw the paper together by discussing the l o g i c a l exten-sion of the r e s u l t s : to describe a means by which the concept of segmented audience based measurement could be incorporated into the decision process concerning the placement of adver-tisements . An advertiser who plans a newspaper campaign i s c l e a r l y not i n a pos i t i o n to tabulate the source data used i n this thesis nor to conduct an analysis such as was done i n Chapters V and VI. Unless, he were an extremely heavy advertiser, i t i s doubtful whether he would have the resources or the i n c l i n a -t i o n to do so. However, i f he were interested i n r a t i o n a l decision-making he would no doubt be interested i n the i n f o r -mation i f i t were available. The conclusion to t h i s study, then, i s directed at the managers of the p r i n t media, the people who s e l l the newspaper as an advertising medium. I t was i n the interests of these people that the source study was conducted from which th i s thesis drew i t s data. The newspapers as an industry have the necessary resources for more sophisticated audience measurement and i t i s strongly i n t h e i r competitive interests with other media to provide accurate information for economic decision-making. Obviously, the p r i n t media managers cannot provide exposure results by section for every conceivable campaign placement open to t h e i r advertising c l i e n t s . However i t would not be d i f f i c u l t y , through a simple extension over Chapter VI, to develop the average expected exposure to the various newspaper sections i n a single issue. For example, the exposure results for each of the six issues could be c a l -culated and then averaged. Then, through simple s t a t i s t i c a l inference, the expected exposure to the entire population could be found. S i m i l a r l y , i t would be possible to calculate the expected net increment to duplicated and unduplicated audience 6 8 . of successive issues of the newspaper. Furthermore, the newspaper could provide the average expected duplication among audience segments. The r e s u l t would be a highly sophisticated breakdown of the newspaper's expected audience on an audience segment basis. The advertiser would then have at his disposal a set of data which i s highly relevant to his campaign objectives. He could test his campaign against the exposure data and have a reasonably sure estimate of the exposure results he can expect to achieve. Of course, exposure does not guarantee advertising effectiveness. As noted before, the effectiveness of an advertising campaign i s very much a r e s u l t of other fac-tors as well as campaign placement. These other factors, such as the qual i t y of the advertisement or the product, are not i d e n t i f i a b l e a p r i o r i by media managers. However the news-papers can s t i l l provide considerable decision f l e x i b i l i t y i n dealing with these e x t e r n a l i t i e s . Data developed on the basis of d i f f e r e n t d e f i n i t i o n s of sections and, more particu-l a r l y , d i f f e r e n t levels of readership enable the advertiser to co-ordinate the placement decision with the campaign objec-t i v e . In short, the advertiser's knowledge concerning the exposure re s u l t s of his campaign i s enhanced considerably. By incorporating expected audience segment exposure into his deci-sion, he i s better able to plan the placement of a campaign a c c o r d i n g t o h i s a d v e r t i s i n g o b j e c t i v e s . T h i s s o p h i s t i c a t i o n i n i n f o r m a t i o n s h o u l d encourage the a d v e r t i s e r t o use t h e newspaper more o f t e n , t o use i t more e f f i c i e n t l y and t o compare i t f a v o u r a b l y w i t h o t h e r media. I f the newspaper today i s l e s s c o m p e t i t i v e and i n c r e a s i n g l y l e s s i m p o r t a n t as an adver-t i s i n g medium, i t i s perhaps because th e i n f o r m a t i o n p r e s e n t l y g e n e r a t e d i s l a g g i n g b e h i n d the modern a d v e r t i s e r ' s d e c i s i o n needs. I n the f u t u r e , b e t t e r i n f o r m a t i o n s h o u l d p r o v i d e b e t t e r r e s u l t s . There are two problems i n the d i r e c t a p p l i c a t i o n o f the segmented aud i e n c e concept w h i c h , a l t h o u g h they were l e f t out of the above d i s c u s s i o n were c o n s i d e r e d a t l e n g t h elsewhere i n the paper. F i r s t , t h e r e i s the problem of c a l c u l a t i n g the u n d u p l i c a t e d audience o f c o m b i n a t i o n s o f newspaper s e c t i o n s and s p e c i f i c a d v e r t i s i n g compaigns. O b v i o u s l y t h e method used i n C hapter V i s not manageable f o r e i t h e r t h e newspaper or the a d v e r t i s e r c o n s i d e r i n g the - i n c r e a s e d number of s e c t i o n s w i t h complete d a t a and t h e c o n s t a n t u p d a t i n g and p r o l i f e r a t i o n * o f p o s s i b l e c o m b i n a t i o n s of i s s u e s and s e c t i o n s . A s i m p l e e s t i m a t i n g p r o c e d u r e f o r u n d u p l i c a t e d a u d i e n c e i s needed and was d i s c u s s e d i n C hapter I I I on r e l a t e d l i t e r a t u r e . T h i s method, u n f o r t u n a t e l y , has n o t been t e s t e d on i n t e r - s e c t i o n d u p l i c a t i o n f o r newspaper d a t a . N e v e r t h e l e s s , any p r a c t i c a l Re: Method used i n C h a p t e r V i s the t h e o r e t i c a l p r o c e d u r e f o r N s e c t i o n s o u t l i n e d i n e q u a t i o n ( 2 ) , Chapter IV, p. 34. 70. application of the segmented audience concept for newspapers w i l l succeed or f a i l over the a b i l i t y to apply this or some closely related net audience estimation procedure. The second problem concerns the trade-off between reach and frequency for a set of advertising campaigns. This trade-off represents the fundamental rela t i o n s h i p for the placement decision. However, i t can only be evaluated at this point by developing a set of campaign alternatives and testing for reach and frequency. There i s no means for deriving the relat i o n s h i p between reach and frequency and incorporating i t into the planning decision before the cam-paigns are designed. The data on exposure to each section do however provide a sound basis for a limited evaluation of this trade-off. The advertiser w i l l be aware that the expo-sure to certain sections i s of no significance to him and w i l l probably be concerned with reach/frequency trade-off for only a r e s t r i c t e d set of alternatives. Problems aside, i t i s the conclusion of this thesis that the segmented audience concept provides a superior means of audience measurement. I t i s e a s i l y foreseeable that a l l manner of supplementary data can be added to the information system once audience segments are established as the approp-r i a t e targets of newspaper advertising campaigns. These might include audience c h a r a c t e r i s t i c s , effectiveness of campaign content and patterns of readership. The only serious obstacle to an immediate application to the advertising decision process i s a simple method of net audience estima-t i o n . However the research of the Agostini t r a d i t i o n i s strongly disposed to the solution of that d i f f i c u l t y . Areas for Further Study 1. The f i r s t area for further study obviously i s a duplication of this study with a stronger orientation to developing a set of data on which advertising decisions can be based.. Now that the a p p l i c a b i l i t y of segmented audience has been v e r i f i e d , the objective should be a complete and workable data base for further research. 2. An interesting but not necessary area for further study would be to examine the overlap or duplication between s p e c i f i c audience segments. For instance, how many readers of the finance section would be duplicated by the sports section; as opposed to the women's section? 3. An extremely important area for examination i s to test the Agostini relationship for inte r - s e c t i o n duplication within a publication. If the relat i o n s h i p i s found to hold then a means of deriving K for newspaper sections must be developed. If the Agostini relationship i s found not to hold true, then further research should be done to discover a workable r e l a -tionship between duplicated and unduplicated audience probably s t i l l using two-by-two section duplication. 4. F i n a l l y , a method must be d e r i v e d t o p r e d i c t , i f p o s s i b l e , a r e l i a b l e r e l a t i o n s h i p between r e a c h and f r e q u e n c y f o r c o m b i n a t i o n s of s e c t i o n s . I t has been demonstrated t h a t b o t h cannot be s i m u l t a n e o u s l y maximized, b u t i s t h e r e a means of e s t i m a t i n g the e f f e c t a change i n one w i l l have on the o t h e r ? Such a method would depend upon o b s e r v a t i o n of the r e l a t i o n s h i p between the v a r i a b l e s o v e r a p e r i o d o f time f o r s p e c i f i c c o m b i n a t i o n s of s e c t i o n s and d e v e l o p i n g d a t a upon e x p e c t e d t r a d e - o f f s as was d e s c r i b e d f o r e x p e c t e d s e c t i o n exposure. Footnotes 1Bogart, L. , "Is i t Time to Discard the Audience Concept? i n Journal of Marketing, Vol. 30 (January, 1966), p. 47. 2 Kuhn, W. , "Net Audience of Vehicle Combinations—m Germany: A New Formula," i n Journal of Advertising Research y Vol. 3 (March, 1963), p. 30. 3 Agostini, J . M., "How to Estimate Unduplicated Audiences i n Journal of Advertising Research, Vol. 1 (March, 1961), p. 11. 4 Ibid., p. 11. ^Ibid., p. 12. ^Ibid., p. 12. 7 Ibid.., p. 13. Q Bower, J . , "Net Audiences of Vehicle Combinations--i n the U.S. and Canada," i n Journal of Advertising Research, Vol. 3, (March, 1963), p. 14. 9 . . . Marc, M., "Net Audiences of Vehicle Combinations—in France," i n Journal of Advertising Research, Vol. 3 (March, 1963), p. 26. 1 0 C a f f y n , J. M. and Sagovsky, M., "Net Audiences of Vehicle Combinations—in B r i t a i n , " i n Journal of Advertising  Research, Vol. 3 (March, 1963), p. 21. ^Kuhn, W. , "Net Audiences of Vehicle Combinations," p. 30. 12 Metheringham, R. A., "Measuring the Net Cumulative Coverage of a Pr i n t Campaign," i n Journal of Advertising  Research, Vol. 4 (December, 1964), p. 23. 13 Hofmans, P., "Measuring the Cumulative Net Coverage of Any Combination of Media," i n Journal of Marketing Research, Vol. 3 (August, 1966), p. 269. Claycamp, J. H. and McClelland, C. W., "Estimating Reach and the Magic of K," i n Journal of Advertising Research, Vol. 8 (June, 1968), p. 44. Ibid., p. 49. BIBLIOGRAPHY Agosti n i , J. M. "How to Estimate Unduplicated Audiences," Journal of Advertising Research. Vol. 1, (March, 1961), pp. 11-14. Bogart, L. "Is i t Time to Discard the Audience Concept," Journal of Marketing. Vol. 30, (January, 1966), pp. 47-54. Bower, J. "Net Audiences of U.S. and Canadian Magazines: Seven Tests of Agostini's Formula," Journal of Adver-t i s i n g Research. Vol. 3, (March, 1963), pp. 13-20. Caffyn, J. M. and Sagovsky, M. "Net Audiences of B r i t i s h Newspapers: A Comparison of the Agostini and Sainsbury Methods," Journal of Advertising Research. Vol. 3, (March, 1963), pp. 21-25. Claycamp, H. J. and McClelland, C. W. "Estimating Reach and the Magic of K," Journal of Advertising Research. Vol. 8, (June, 1968), pp. 44-51. Hofmans, P. "Measuring the Cumulative Net Coverage of Any Combination of Media," Journal of Marketing Research. Vol. 3, (August, 1966), pp. 269-278. Kuhn, W. "Net Audiences of German Magazines: A New Formula," Journal of Advertising Research. Vol. 3, (March, 1963), pp. 30-33. Marc, M. "Net Audiences of French Business Papers: Agostini's Formula Applied to Special Markets," Journal of Adver-t i s i n g Research. Vol. 3, (March, 1963), pp. 26-29. Metheringham, R. A. "Measuring the Net Cumulative Coverage of a P r i n t Campaign," Journal of Advertising Research. Vol. 4, (December, 1964), pp. 23-28. A P P E N D I C E S 76. APPENDIX A Ta b l e I I . R e a d e r s h i p l e v e l = .1 Campaigns 1 (A) 1 (B) 2 (A) 2 (B) 3 (A) 3 (B) Aggregate D u p l i c a t e d A u d i e n c e 1194 1030 2140 2439 3225 3422 3758 U n d u p l i c a t e d A u dience 333 325 358 396 373 373 374 Frequency o f Exposure 3.59 3.17 5.98 6.61 8.65 9.17 10.05 R e a d e r s h i p l e v e l = .25 Campaigns 1 (A) 1 (B) 2 (A) 2 (B) 3 (A) 3 (B) Aggregate D u p l i c a t e d A u d i e n c e 908 867 1771 2085 2618 2825 3096 U n d u p l i c a t e d A u d i e n c e 274 305 344 360 356 362 364 Frequency o f Exposure 3.32 2.84 5.15 5.79 7.35 7.80 8.51 R e a d e r s h i p l e v e l = .75 Campaigns 1 (A) 1(B) 2(A) 2(B) 3(A) 3(B) Aggregate D u p l i c a t e d A u d i e n c e 291 396 693 878 1029 1114 1233 U n d u p l i c a t e d Audience 144 203 239 259 252 269 272 Frequency o f Exposure 2.02 1.95 2.90 3.39 4.08 4.14 4.53 77. APPENDIX B Table III Readership l e v e l = .1 Sports Finance Women's Entertmt. Aggregate Duplicated Audience 1194 831 1200 533 3758 Unduplicated Audience 333 278 318 316 374 Frequency of Exposure 3.59 2.99 3.78 1.69 10.05 Readership l e v e l = .25 Sports Finance Women's Entertmt. Aggregate Duplicated Audience 908 687 1023 478 3096 Unduplicated Audience 274 238 280 294 364 Frequency of Exposure 3.32 2.89 5.66 1.63 8.51 Readership l e v e l - .75 Sports Finance Women's Entertmt. ; Aggregate Duplicated Audience 291 268 470 204 1233 Unduplicated Audience 144 114 158 141 272 Frequency of Exposure 2.02 2.35 2.97 1.45 4.53 APPENDIX C Table IV 78, Readership l e v e l = . 1 Campaigns Duplicated Audience Unduplicated Audience Frequency of Exposure 1194 831 333 278 3.59 2.99 Readership l e v e l = .25 Campaigns 1 2 Duplicated Audience Unduplicated Audience Frequency of Exposure 908 687 274 238 3.32 2.89 Readership l e v e l = .75 Campaigns 1 2 Duplicated Audience Unduplicated Audience Frequency of Exposure 291 268 144 114 2.02 2.35 1200 318 1023 280 3. 66 470 158 4 1057 327 3.78 3.23 283 285 144 5 1170 345 3.39 839 1014 331 2.97 3.06 2.97 1.98 409 207 1.98 79 APPENDIX D Table V Readership l e v e l = .1 Sports 1 2 3 4 5 DA 1194 - - 609 394 UDA 333 - - 300 259 F 3.59 - - 2.03 1.52 Finance 1 2 3 4 5 DA - 831 - 448 132 UDA - 278 - 236 132 F - 2.99 - 1.90 1.00 Women's 1 2 3 4 5 DA - - 1200 - 383 UDA - - 318 - 240 F - -; 3.78 - 1.59 Entertainment 1 2 3 4_ 5 DA - 261 UDA - - - " 2 61 F - - 1.00 80. Readership l e v e l = .25 Sports 1 2 3 4 5 DA 908 - - 496 346 UDA 274 - - 248 224 F 3.32 - - 2.00 1.54 Finance 1 2 3 4 5 DA - 687 - 343 107 UDA - 238 - 181 107 F - 2.89 - 1.89 1.00 Women's 1 2 3 4 5 DA - - 1023 - 328 UDA - - 280 - 204 F - -- 3.66 - 1.61 Entertainment 1 2 2 - 1 I DA 233 UDA - - - 233 F - 1.00 81. R e a d e r s h i p l e v e l -- .75 S p o r t s 1 2 3 4 5 DA 291 - - 179 140 UDA 144 - - 126 118 F 2.02 - - 1.42 1.19 F i n a n c e 1 2 3 4 5 DA - 268 - 106 35 UDA - 114 - 63 35 F - 2.35- - 1.68 1.00 Women's 1 2 3 4 5 DA - - 470 - 138 UDA - - 15 8 - 9 8 F - 2.97 - 1.41 E n t e r t a i n m e n t 1 2 3 4 5 DA 96 UDA - - - - 9 6 F _ _ _ _ i .oO 82. APPENDIX E Table VI Readership l e v e l - .1 Duplicated audience Net Reach DA (sports) = 609 DA (finance) =. 448 DA (campaign) = 1057 UDA (sports) = 300 UDA (finance) = 236 Duplication of readers between sports and finance = 209 UDA (campaign) = (300 + 236)-209 = 327 Frequency of Exposure F (sports) F (finance) F (campaign) 609 300 448 236 1057 327 = 2.03 = 1.90 = 3.23 83. Readership l e v e l = .25 Duplicated Audience DA (sports) = 496 DA (finance) = 343 DA (campaign) = 83 9 Net Reach UDA (sports) = 248 UDA (finance) =. 181 Duplication of readers between sports and finance = 146 UDA (campaign) = (248 + 181)- 146 283 Frequency of Exposure 496 F (sports) = — ° = 2.00 3 43 F (finance) = ~ - = 2.89 a 3 q F (campaign) = = 2.97 248 4181 8 9283 84. Readership l e v e l = .75  Duplicated Audience DA (sports) = 179 DA (finance) = 106 DA (campaign) = 2 85 Net Reach UDA (sports) = 126 UDA (finance) = 63 Duplication of readers between sports and finance = 45 UDA (campaign) = (126 + 63)- 45 = 144 Frequency of Exposure F (sports) = ^yf- = 1.42 F (finance) - = 1.61 285 F (campaign) = -. . . = 1.9 8 126 10( 63 !144 85. APPENDIX F Table VII Readership l e v e l = .1 Campaigns Duplicated Audience Net Reach Frequency of Exposure 1. 1194 333 3. 59 2. 831 378 2. 99 3. 1200 318 3.78 4. 1057 327 3.23 .5. 1089 335 3.25 6. 1051 325 3.23 7. 1016 329 3. 08 8. 1228 337 3.64 9. 1031 318 3. 24 10. 1000 315 3.18 11. 1170 345 3.39 12. 1195 338 3.54 13 . 1087 339 3.20 14. 1019 339 3.01 15. 1140 339 3.36 16. 1258 347 * 3.63 86. Readership l e v e l = .25 Duplicated Frequency of Campaigns Audience Net Reach Exposure 1. 908 274 3.32 2. 687 238 2.89 3. 1023 280 3.66 4. 839 283 2.97 * 5. 912 316 2.89 6. 859 299 2.87 7. 856 309 2.77 8. 1044 328 3.18 9. 837 281 2.98 10. 873 293 2.98 11. 1014 331 * 3.06 12. 1039 331 * 3.14 13. 912 308 2.96 14. 371 318 2.74 15. 998 325 3.07 16. 1041 329 3.16 87. Readership l e v e l = .75 impaigns Duplicated Audience Net Reach Frequency of Exposure 1. 291 144 2.02 2. 268 114 2.35 3. 470 158 2.97 * 4. 285 144 1.98 5. 342 190 1. 80 6. 351 165 2.13 7. 383 199 1.93 8. 406 180 2.26 9. 321 150 2.14 10. 417 174 2.40 11. 409 207 1.98 12. 438 210 2.09 13. 358 156 2.30 14. 288 145 1.99 15. 461 218 * 2.12 16. 399 210 1.90 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0101974/manifest

Comment

Related Items