~ — .L-A .— .1'- —- nan: 1‘ . Ar": n.1, J .. .._.. A41 " r 71:. 1: .1- -r 7.: Arr L’ .4 'v 4:5".-. ‘ .1 .r u“: n;-:.— A ._ :‘r A“: . _“.:,:::_r":;x “1131321415 4 so ,- r‘z‘ ~ {‘7' "+4 A. .J‘ .3}: . .,(‘,.“ 3n... 5;: :1. £491.75 ”gig 1-“ '4. {-2.41 g; .4 J‘::“ git; L 34" .2 ‘4‘ 15 441-31 " 3 u”;$::.:'=£:£?£f ~75 57¢. 3.2:; m”? Ad's... ' ~ g 5 2131193..- 1:; .3." ~1- ,. , ‘ ’ . . ..2'.'1‘§11:":‘§‘:' a , 31‘3"" ‘17-..1. :, 111;” Win“! ,. 1:“ ~3"*":' 41.2. ‘1‘ f 7 «m: “of“ 111:. : ~ ~ ~19: . 1.1.1111! iM11111“ 51...:11 ‘61: 1". 6.! 'ip: [1.11.143 "‘1: . }7" .~ ‘1‘“ .;='1',“:;‘ ‘1“1‘3’ 1&1: ‘1“ ‘t"“' : .. '1... . ( @1111?" t 1 '1‘ 11:“ “1"“ I‘1'u' ‘1 l. I; a V. 1 ‘1. {Ex 11' ""“W “1.511%: ' 3:15: ‘2‘ ‘1'." 1&wa "E313“ 1‘ ll. .11: 12"“ $161“; 1‘: .2 . “ ’ . ‘3 . . L ' " 1:: 11': 3“ . 11‘?“ ..r "'1." 141'} . 11:11; 1 «11-1182 . l 1“)" ' A“ $54315" ‘i'ii‘ik "K 1:.‘11: 17:11:11.3 {11:11 ‘1‘ 3311;- :2 1 1;": 52"" a '1: i: ' 1 $1?! ‘2 "‘9 '1‘ f -. . v‘ 7 £11 “’7“ 3'“ :3: .1‘114511‘11‘1 4"?“ 1*“ i .‘ .13'11,‘ Egggtt‘. . 11:1- :1. : ““51 {13: 0,3118? 511‘? 1:11: I: 1: 'flan‘tgg‘fm} . :1 "d 1 :1“, 4533‘ "\ ‘1 1,1 "1 95"? 1' E113.‘ ‘ ‘ Linagi‘g-b ', A xtEi': ‘4‘ I”? M ‘1. A. TH." , A .r x #3,. ' .4... ”34;" =‘ 3‘4. ‘1 2' r: .1 L r “j." 41.... ”Ls” *4”! :- — '9 ”31:3,?“ 7"“ 5‘ (21.13:_ P J 5 fl ' " m-.. A: .r«f*"' Err-v 1" r 4:" 4 11.1?” fi ‘1 - Vt: 5‘3 —=‘ 4 f’ A “ -“ *' Hz.“ '3': fl. - i"? “:11: 1?]. 4 n , . Z '1“ , 1‘ 1 :ATQ‘QXEEQEA‘ 31, r 6‘ ‘ 1"» 1 ’7’: 2‘ V I I :‘z’: .. 1 21”“.' ' ‘ J a... m ‘*‘ ' "‘2" "’1 ‘ ‘ v. -' «‘1' 15:11.17." " .‘2‘. 2-- W ,. L»- F": :3}? i:— .‘ 11.! «‘3‘ . “ $5 #7 A: “a wzfiffiz: T "Y. ‘ .55.? ‘22:? ‘ .7!“ “A 2211: L‘ ‘1 = $211 ‘§._‘..v ' ‘1? z‘ 1"}; ’1'.“ 1‘11ij {‘1 ‘ "1‘ £161: 1': “55"" ' ' t". "‘ ‘ "ké ' «a. .11.} .1” ‘11"th 11‘: fl“: . 1 311:"‘7‘114 1...,» - :16“; -. 11‘" ’ 1.: . 4: .1“ ‘ Raj-1" 1:111; . «11:11:: . 5&1' :‘4'. 1% .1. 4:: :1 :3: :1, . . ‘11:: $3“ 131,11: if, ‘4 Vita; ‘. :1‘1’1.‘ “I: .' {.1} '1. 3.1.5111 ‘ «5“. :31: , . “31% «b 19.2 3‘- ”1": £333. .‘1'1 ‘11:. 1, 4‘: 1.13:}: “23.1.11. I.” ‘1. ‘ ‘3 211:."1‘n. :4}: “1 1- 1“"? "l lllllllll llllllllllllIl'flllllfllll L 3 1293 010§8 0734 ass: 3 f , some g. ._:,‘l “organ mac» University w are“. l This is to certify that the dissertation entitled THE IMPACT OF COUPONING ON BRAND IMAGE OVER REPEATED EXPOSURE: , AN EXPERIMENTAL ANALYSIS ‘ presented by Tamara Suzanne Brezen has been accepted towards fulfillment of the requirements for Ph.D. . Mass Media \ degree in / Major professor Date / “5/ A ’ MS U is an Affirmative Action/Equal Opporluniry lnstilun‘on 0« 12771 MSU LIBRARIES RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wiI] be charged if book is returned after the date stamped be10w. 6 9 ”flag fikzfijfifl em M MW C I. ~ we r a. , THE IMPACT OF COUPONING 0N BRAND IMAGE OVER REPEATED EXPOSURE: AN EXPERIMENTAL ANALYSIS By Tenara Suzanne Brezen l A DISSERTATION Submitted To Michigan State University in partial fulfill-eat of the requirements i for the degree of , DOCTOR OF PHILOSOPHY r" l College of Communication Arts and Sciences——Mass Media 1986 Copyright by TAMARA SUZANNE BREZEN 1986 ABSTRACT THE IMPACT OF COUPONING 0N BRAND IMAGE OVER REPEATED EXPOSURE: AN EXPERIMENTAL ANALYSIS By Tamara Suzanne Brezen Sales promotion now commands over 60% of the marketing dollars spent by package goods companies, yet almost no published research is available about the long—term implications of its use and, therefore, little planning precedes its implementation. Couponing represents the majority of these promotion dollars directed toward consumers and there is widespread concern over whether this increasing use of couponing may damage or weaken 8 brand’s image and future franchisement. This fear is partly grounded in case history analyses of sales trends which show sales declining with increasing use of sales promotion relative to advertising. However, the aggregate and uncontrolled nature of this data lends it imprecise for planning purposes and insensitive to the influence sales promotion may be exercising on a brand’s communication process prior to purchase. This is an exploratory study of such effects given a learning and information processing perspective. A laboratory experiment was designed to explore the impact of print advertisements with coupons (sales promotion) versus identical ads without coupons over one, two, and three exposures. Eight product categories, primarily health and beauty aids, were selected for study among a high-usage target audience—-393 women, 18 to 34 years of age. Chi square analyses on brand and advertising recall, and analyses of variance on brand evaluations, purchase intentions and cognitive elaborations were conducted to test for differences across repetition or promotion conditions. Dealing, brand loyalty, and purchase frequency-—covariates in this study—- appeared to be product specific rather than general consumer traits and these product category differences necessitated separate category analyses. Repetition was found to generally increase recall levels as learning theory would predict, but significant effects on higher-level attitudinal measures were not found. There was evidence that coupons increase perceived brand value for some brands and diminish the cognitive response to a brand and its advertising. Coupons, in accordance with distraction theory, appeared to distract from cognitive responses, hurting brands/advertising with otherwise positive response while, aiding products or advertising eliciting negative response. In this analysis, the cognitive response measures provided more insight into the mediating effects of sales promotion than the higher—level outcome measures such as attitudes and purchase intentions which are typically used. Most importantly, this study showed that sales promotion effects vary by product category. ACKNOWLEDGEMENTS I would like to express my gratitude to Dr. Martin P. Block, my dissertation advisor, for his patience, guidance, and friendship throughout this project and the last three years. Under his direction, I have learned a tremendous amount about advertising and research. Without his encouragement, I may never have pursued this degree. I would also like to thank the other members of my committee: Dr. Bruce Vanden Bergh, Dr. Keith Adler, and Dr. John Abel, for their contributions to the development of this study. I am deeply indebted to the Council of Sales Promotion Agencies which provided both financial support to begin the research and the incentive to finish! To my family, I owe a special thanks for the patience, understanding and emotional support that was always there when I needed it most. Last, but not least, I thank Abigail for providing the sanity and comfort that only a puppy can give. TABLE OF CONTENTS Page Chapter I. INTRODUCTION . . . . . . . . . . . . . . . . . . 1 The Growth of Sales Promotion . . . . . . . . . 1 Past Research Directions . . . . . . . . . . . 4 Price Promotions and Brand Attitudes . . . . . . 11 Study Purpose . . . . . . . . . . 15 Organization of the Dissertation . . . . . . . . 15 II. REVIEW OF THE LITERATURE . . . . . . . . . . . . 17 Defining Sales Promotion . . . . . . . . 18 Basic Concepts in Sales Promotion Research . . . 25 Brand Image . . . . . . . . . . . . . . . 25 t The effects of sales promotion on brand image . . . . . . . . . . . . . 27 Measuring brand image . . . . . . . . . 29 Brand Loyalty . . . . . . . . . . . . 31 Brand loyalty correlates . . . . . . . . 34 Measuring brand loyalty . . . . . . . . 36 Deal Proneness . . . . . . . . . . . . . 39 Measuring deal proneness . . . . . . . . 42 Theoretical Perspectives for Sales Promotion . . 45 Cognitive Dissonance and Attribution Theory . . . . . . . . . . . 45 Cognitive dissonance theory . . . . . . 45 Attribution theory . . . . . . . . . . . 49 Self perception theory . . . . . . . . . 51 Price Perception Theories . . . . . . . . . 54 Adaptation theory . . . . . . 55 Assimilation- -contrast and Weber’ s Law . 56 Learning Theories . . . . . . . . . . . . . 58 Behavioral learning theories . . . . . . 60 Classical conditioning . . . . 60 Operant conditioning and shaping . . 61 Cognitive learning theory . . . . . . . 65 Repetition and cognitive learning . . . . . . . . . 67 Multi- attribute models . . . . . . . 72 Information Processing . . . . . . . . . 73 Screening and encoding sensory input . . 76 Rehearsal, storage, and retrieval . . . 78 Cognitive responses . . . . . . . . . 82 Repetition and cognitive response . . . . . . . . . . 85 Distraction theory . . . . . . . . . 87 Information overload . . . . . . . . 88 Summary: Putting it all in Perspective . . 89 iii III. Iv. ’Research Design . . . . . TABLE OF CONTENTS (Continued) Exposure Frequency . . . . . . . . . General Repetition Effects . . . . . . . Exposure Control . . . . . . . . . . . . Research Questions and Hypotheses . . . . VMETEODOLOGY . . . . . . . . . . . . . . . . Procedure . . Disguise of the Experiment Sample . . . . . . . . . . . . . . Selection and Preparation of Stimuli Materials . . . . . . . . . . . . . . . . Operationalization of Experimental Variables Independent Variables . . . . . . . . . Repetition . . . . . . . . Promotion manipulation . . Dependent Variables . . . . . Brand recall . . . . . Advertising related recall 9 s e e s and coupon mention . . . Cognitive response . Brand evaluation . . . . Purchase intent . . . Other Experimental Variables Product usage . . . . . Brand loyalty . . . . . . Coupon usage and deal proneness Pretesting of the Procedure and Instrumen Data Analysis Procedures . . . . . . . . Limitations . . . . . . . . . . . . Design . . . . . . Variable Operationalizations . . . a o e e e s e e o e s o o s e e s o o ff 0 s s s e s s s s s 0 ANALYSIS OF EXPERIMENTAL RESULTS Sample Characteristics . Demographics . . . . Media Usage . . . . Product Usage . . . Brand Loyalty . . Deal Proneness . . Data Recoding and Analysis Proc Individual Category Analyses Hair Remover . . . Nail Polish Remover . Mascara . . . . . . . d r Headache Remedies . Powdered Drink Mixes Sanitary Napkins . . s o s s e s s o s e s s s m e e e o e s m c o e e m s m e s s o s s o s o e o e o e s s e e e s e a o s s O I e s s e s s s e s I e s s s e o e 0 iv e s e s e s o s e e e m s e s s s o e e e e s o o s e e s s e s e s I s s o e m e e s s v s o e 104 106 108 111 112 115 115 115 116 117 118 119 121 126 126 126 127 131 132 134 135 135 136 139 139 139 141 141 142 144 146 147 148 152 155 159 162 165 TABLE OF CONTENTS (Continued) Deodorant . . . . . . . . . . . . . . Pace and Body Soap . . . . . . . . . . . V. SUMMARY AND CONCLUSIONS . . . . . . . . . . Summary . . . . . . . . . Brand and Advertising Recall . Brand Evaluation . . . . . . . . Value and Purchase Intentions . Cognitive Responses . . . . . . Coupon Recall . . . . . . . . . . Advertising Versus Sales Promotion Ve Both . . . . . . . . . . . . . Brand Advertising Recall over Repetition Brand Evaluation, Brand Value, Purchase Intent, and Cognitive Response over Repetition . . . . . . . . . . . . Conclusions . . . . . . . . . . . . Future Research Directions . . . . In the Laboratory . . . . . . In the Field . . . . . . . . essru APPENDICES A. Campaign Advertisements . . . . . . . . . . B. Test Advertisements . . . . . . . . . . . . C. Data Collection Instrument . . . . . . . . . D. Cognitive Response Category Descriptions . B. Analysis of Variance Summary Tables for Advertising Versus Sales Promotion . . . F. Analysis of Variance Summary Tables for Advertising and Sales Promotion Mixed . BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . O O 0 O O O LIST OF TABLES Table Page 1.1 Average Redemption Rates According to Delivery Method . . . . . . . . . . . . . . 11 2.1 Brand Loyalty Measures in Brown Study . . . . . 37 2.2 Brand Loyalty Measures . . . . . . . . . . . . . 40 2.3 Theoretical Perspectives for Sales Promotion Effects . . . . . . . . . . . . . . . . . . 92 2.4 Approaches in Repetition Research . . . . . . . 98 3.1 Experimental Conditions . . . . . . . . . . . . 105 3.2 Experimental Cell Sizes . . . . . . . . . . . . 111 3.3 Test Brands and Advertisement Characteristics . . . . . . . . . . . . . . 114 3.4 Number of Exposures Tested in Past Frequency Research . . . . . . . . . . . . . 116 3.5 Aided Recall Coding Examples . . . . . . . . . . 120 3.6 Cognitive Response Categories . . . . . . . . . 122 3.7 Cognitive Response Coding Examples 124 a s s o e s 4.1 Sample Demographic Characteristics . . . . . . . 140 4.2 Purchasing Patterns, Brand Loyalty, and Coupon Usage of Sample . . . . . . . . . . . . . . 143 4.3 General Brand Loyal and Dealing Behavior . . . . 144 4.4 Average Coupon Redemption Per Month . . . . . . 145 4.5 Hair Remover Effects . . . . . . . . . . . . . . 149 4.6 Hair Remover Cognitive Response . . . . . . . . 151 4.7 Hair Remover Coupon Mention by Deal Proneness . 152 4.8 Nail Polish Remover Effects . . . . . . . . . . 153 4.9 Nail Polish Remover Cognitive Response . . . . . 154 4.10 Nail Polish Remover Coupon Mention by Deal Proneness . . . . . . . . . . . . . . . . . 155 4.11 Mascara Effects . . . . . . . . . . . . . . . . 157 4.12 4.13 4.14 4.15 4.16 4.17 4.18 4.19 4.20 4.21 4.22 4.23 4.24 4.25 4.26 4.27 LIST OF TABLES (Continued) Mascara Cognitive Response . . . . . . . . . . . Mascara Coupon Mention by Deal Proneness . . . . Headache Remedy Effects . . . . . . . . . . . . Headache Remedy Cognitive Response . . . . . . . Headache Remedy Coupon Mention by Deal Proneness Drink Mix Effects . . . . . . . . . . . . . . . Drink Mix Cognitive Response . . . . . . . . . . Powdered Drink Mix Coupon Mention by Deal Proneness . . . . . . . . . . . . . . . . . Sanitary Napkin Effects . . . . . . . . . . . . Sanitary Napkin Cognitive Response . . . . . . . Sanitary Napkin Coupon Mention by Deal Proneness Deodorant Effects . . . . . . . . . . . . . . . Deodorant Cognitive Response . . . . . . . . . . Deodorant Coupon Mention by Deal Proneness . . . Soap Effects . . . . . . . . . . . . . . . . . . Soap Cognitive Response . . . . . . . . . . . . Soap Coupon Mention by Deal Proneness . . . . . Frequencies and Chi—Square Summaries for Brand Recall Across all Product Categories Frequencies and Chi-Square Summaries for Advertising-Related Recall Across all Product Categories . . . . . . . . . . Frequencies and Chi-Square Summaries for Coupon Recall According to Deal Proneness Categories . . . . . . . . . . . . . . . . . Analysis of Variance Summary Tables for Brand Value Across Product Categories . . . . . . Analysis of Variance Summary Tables for Brand Evaluation Across Product Categories . . . Page 158 158 160 161 161 163 164 165 166 167 168 170 171 171 173 174 174 220 224 228 231 234 F02 F.3 LIST OF TABLES (Continued) Analysis of Variance Summary Tables for Brand Purchase Intent Across Product Categories . . . . . . . . . . . . . . . Analysis of Variance Summary Tables for Overall Cognitive Response Across Product Categories . . . . . . . . . . . . . . . . . Analysis of Variance Summary Tables for Product- Related Cognitive Response Across Categories . . . . . . . . . . . . . . . . . Analysis of Variance Summary Tables for Advertising-Related Cognitive Response . . . Frequencies and Chi-Square Summaries for Brand Recall by Product Category . . . . . . Frequencies and Chi-Square Summaries for Advertising-Related Recall by Product category 0 O O O O O 0 0 O O O O O 0 O O 0 0 Analysis of Variance Summary Tables for Brand Value by Product Category . . . . . . . . Analysis of Variance Summary Tables for Brand Evaluation by Product Category . . . . . . . Analysis of Variance Summary Tables for Brand Purchase Intent by Product Category . . . . Analysis of Variance Summary Tables for Overall Cognitive Response by Product Category . . . Page 237 240 242 244 246 249 252 254 256 258 Figure 1.1 1.2 LIST OF FIGURES Growth of Advertising and Sales Promotion, 1969—1981 e e a s e e s o n e e s a e e 0 Comparison of Advertising and Promotion Expenditures and Sales for Ribena . . . . Sales of a Promoted Brand Versus an Unpromoted Brand 0 O O C O O O O O O O O O O O O O O Allocation of Dollars to Sales Promotion Techniques . . . . . . . . . . . . . . . . Growth of Couponing, 1980-1984 . . . . . . . The Three Modes of Marketing . . . . . . . . The Sales Promotion—Advertising Continuum . . Latitudes of Acceptance in Price Perceptions . Classic Learning Curve . . . . . . . . . . . Two Factor Theory of Affect Toward a Repeated Stimulus . . . . . . . . . . . . . . . . . Repeated Measures Design Across Eight Brands . ix 0 Page 12 13 22 24 57 58 69 106 CHAPTER I INTRODUCTION The Growth of Sales Promotion Advertising is critical in consumer product marketing, working to differentiate a branded product from its competi— tors. Especially for products which have many hidden benefits yet few tangible assets, or for parity products, advertising is a way to create a distinct image for a brand which can build a stable, loyal customer base and protect that brand franchise against competitive inroads. Sales promotional activities, however, such as couponing, sampling, refunding, and on-pack price cutting, as well as the use of sweepstakes, premiums, bonus packs, and trade incentives have gained marketing importance. The popularity of sales promotion has been especially evident in the past two decades. Since 1969, sales promotion expenditures have been growing steadily and at a rate faster than that for advertising expenditures as Figure 1.1 demonstrates. Package goods sales promotions totaled $30 million in 1984 and these dollars account for 60 percent of the marketing dollars spent by package goods companies (Meyers, 1985). Reasons for this growth can be traced back to the recessionary periods of the seventies when consumers were especially budget-conscious and vulnerable to price NN NN NNN N SALES PROMOTION t\\\‘ ADVERTISING Fig. 1.1 The Growth of Advertising and Sales Promotion 1969-1981 promotions. During this period, too, new brands and new product introductions were deluging the market, and the demand for retail shelf space was greater than the supply. Dealing to the trade increased in order to secure the necessary facings on the limited shelf space. Advertising costs soared with inflation and sales promotion was recognized as a more flexible and cost—efficient substitute when dollars ran short. In addition, the successful short- term results from aggressive competitive promotional activity eroded management’s resistance to promoting (Strang, 1976; Petersen, 1980). Throughout the inflationary and intensely competitive times, sales promotion tactics directed to the trade have assisted manufacturers in acquiring retailer shelf space and support. To the consumer, sales promotion helped to induce trial by non-users who might be converted to regular purcha— sers, to permanently increase usage by regular users, or at the very least, to load present customers with the product so as to preempt competitive actions which might promote brand switching. As a result of the industry’s continued and growing use of sales promotion, however, these temporary incentives have lost their uniqueness. Marketers have been forced by compe- tition into retaliatory price promoting to both retailers and consumers to maintain market competitiveness. Retailers have since become addicted to the trade deals and, as they become more cognizant of the power they wield in granting shelf space, aisle displays, or special weekly feature display advertising, retailers are also becoming more sophisticated and demanding in terms of frequency, timing and terms of the promotion (Meyers, 1985). In some instances, retailers have stopped promotions from running, such as in—store giveaways, which a manufacturer had planned (Christopher, 1971). Moreover, what has emerged from the recessionary economy and intense promotional activity of the past decade is a consumer who is more susceptible to price promotion than ever before and "deal prone." This has meant continued and heavier discounting to the consumer through couponing and price—offs. Nearly 180 billion consumer coupons were dropped in 1984, up from 132 billion in 1983 (Vincent, 1984; Meyers, 1985), according to A.C. Nielsen. Even retailers, while demanding trade support, have begun to express some anxiety over the sheer number of consumer promotions ("Coupon Patterns," 1985). Companies are left struggling to maintain brand differ— entiation while these sales promotion efforts create price parity among brands, reduce margins, and threaten to cripple company profits. The fast food industry and the cola manufacturers are currently in this predicament from over promoting and, as one soft drink executive has admitted, the industry is trying new strategies to "get the price monkey off its back" ("Price Junkies," 1984, p.91). Past Research Directions Despite the anxiety advertisers or retailers have over the increasing use of sales promotion, consumer promotions are predicted to continue to expand both in dollars and sophistication in the future (Strazewski, 1986). Therefore, the promotional trap in which many marketers find themselves indicates an obvious need to strategically plan the advertising and promotion program carefully. Unfortunately, this is not the case currently in most companies. Strang’s study of promotion in the United States (1976) and Christopher’s study of the sales promotion industry in Britain (1971) reveal strikingly similar results: very little pretesting and planning precedes implementation of promotions and few companies systematically evaluate either the short term or long term effectiveness of a sales promotion. Given the amount of money spent in sales promo— tion today, it is amazing that so little formal research on sales promotion effects or on the interaction of sales promotion with advertising has been conducted. The published literature that is available on the topic has been focused on the macro~level sales effects and profits of promotion. On one hand, papers which describe sales promotion models based on assumptions about such variables as consumer and trade dealing propensity and purchase behavior can be found. These models attempt to simulate and predict market sales response and profit maximization for sales promotion. 0n the other hand, one can find what are essentially isolated case histories of a brand’s market performance after use of sales promotions. Models to explore the profitability of one or more sales promotions in a brand’s promotional mix have been developed by several authors. For instance, Blattberg, Eppen, and Lieberman (1981) formulated an inventory control model which assume retailers deal to pass along holding costs to consumers willing to stockpile. McAlister (1983) presented a model to predict market response to sales promotions accor- ding to industry competitive structure and consumer deal proneness. These models, while proving useful in the planning stages of a promotion at a time when little empiri— cal data is available for decision making, do not incorporate assumptions about or predict implications of sales promotion on brand image. The empirical evidence that is available is sales and larket share oriented and both the aggregate nature of these Ieasures and the total lack of control over other extraneous marketing and competitive variables, lends it somewhat invalid for precise planning purp0585~ For example, Strang (1975) has reported several case histories where long—term sales or market share declines have Paralleled increases in a brand’s sales promotion expenditures as a proportion of the promotional budget. Ribena and Lucozade, two different British drink products which each dominated their respective markets, encountered sales declines when .8195 promotion increased relative to advertising and experienced sales increases when advertising dollars were increased at the expense 0f 531°“ pr°'°ti°n' Figure 1.2 shows this advertising/promotion relationship Clearly for one of the two products, Ribena. Years I 2 3 4 5 6 Advertisin 120—- ,’ g 110d 100-4 Sales 90 _ Index RH 30 _ 70 -‘ Promotion 60 _. 50 _ Fig. 1.2 Comparison of Advertising and Promotion Expenditures and Sales for Ribena A second case history involving three competitive brands within a disguised consumer good category, reports similar findings. Cutbacks in advertising for the three brands combined during a four year period resulted in a drop in advertising’s proportionate dollars from 63 to 22 percent. Although sales previously had been growing at twice the population rate, the total share of market remained relatively flat during this period, falling slightly from 553 in 1969 to 54x in 1972. Profits for the three brands fell 36* in the first two years alone. The brand which fared the best in market share and profits maintained a higher propor- tion of advertising relative to sales promotion, while the most promotion-oriented brand of the three was hardest hit with sales and profit declines. The results of both these studies are based on aggregate advertising to sales promotion dollar ratios which are not sensitive to differences in sales promotion techniques, creative executional strategies, or different cost structures for advertising and promotion. Furthermore, expenditures mask the relative impact to individuals with respect to reach and frequency of impressions. Since so many other variables are left uncontrolled in this type of longitudinal one—shot case analysis, the observable declines in sales cannot be positively linked to the proportion of dollars spent on sales promotion relative to advertising. Growing evidence from available short term analyses, again on a case history basis, also suggests that sales promotion will not increase brand market shares or brand franchises as expected, and hint that these incentives may mortgage a brand’s future. Clayton (1975) analyzed Nielsen data trends which confirm that all too often, instead of market share increases, sales share often returns to prior pre-promotion levels or decreased levels. A confidential study made available to Sales Promotion Monitor, demonstrates this phenomenon. As Figure 1.3 shows, sales for Brand A increased dramatically during the promotion period but dipped to lower- than-usual levels afterward without affecting Brand B’s market share. This means that the promotion attracted deal prone persons, brand switchers, and regular users who took advantage of the price reduction but later returned to regular usage levels. The promotion does not appear to have stolen users away from the competition. Since the brand franchise was not strengthened, promotions merely diminished the manufacturer’s profits in this example. 80.000 - Units 70.000 _. 60.000 -- 50.000 -- 40.000 ~- 30,000 - 20.000 - BRAND A __ ------ / \-- BRAND B -—-- Fig. 1.3 Sales of a Promoted Versus an Unpromoted Brand. Source: "Discounting Price." Sales Promotion Monitor, 2, 7 (July 1984): 52. Brown (1974) surveyed 6000 households and also concluded that sales promotion efforts, while increasing short term sales response, do not yield permanent market share gains. He noticed, however, that brand loyal persons tended not to deal but repurchased at a higher rate and price conscious people responded initially to the promotion but did not repurchase once the price advantage was removed. An analysis of scanner data by Block and Totten (forthcoming) corroborates this finding, further revealing that brand switchers are responsible for approximately 80% of the purchases made during a promotion period. In heavily promoted categories like colas, there are consumers who are "cherry pickers" and only buy when a brand is on deal——even 10 those loyal to a brand will wait to buy until their favorite brand is promoted and then will stockpile until it is promoted again. A major package goods company claims that over 75x of all coupons redeemed are submitted by consumers who already use the couponed brand (Meyers, 1985). Since couponing and price promotions have been used primarily with the belief that inducing new trial by new customers would lead to the addition of a number who would remain loyal, the reality of reports such as these have no doubt been disturbing and probably have influenced the change in strategy occurring now in the industry. Increasingly, consumer couponing is being used as a merchandising event to secure trade cooperation, a switch from a pull to a push strategy. This shift is noticeable in the delivery method trends for coupons (Vincent, 1984). Although coupons are delivered via many vehicles, print delivered coupons and free standing inserts (FSI’s) are becoming more popular among manufacturers than direct mail coupons because they are less costly. This reduced cost reflects not only the elimination of postage, but also the lower-than-average redemption rates. Magazines, newspapers, and Sunday supplements account for 82% of coupon delivery and are redeemed at 1/2 and 1/4 of the rate of direct mail, respectively ("Coupon Patterns," 1985, P. 11). Table 1.1 shows the average redemption rates for each medium. Couponing as a merchandising strategy operates on different economics; lower redemption rates are desirable because the objectives are to increase brand exposure by ll pushing the product into the distribution and onto the shelves. Inevitably, then, many consumers, both loyal and non-loyal, will see frequent couponing for their brands, intentionally positioned to minimize redemption. From this perspective, the inclusion of price or a coupon in advertising as an advertising component poses new concerns. What is the nature of the influence which sales promotions exercise on a brand’s communication process; and how does this affect brand attitudes or brand image? TABLE 1.1 AVERAGE REDEMPTION RATES ACCORDING TO DELIVERY METHOD Avg. Redemption Medium Rate Percentage Magazine Daily newspaper Sunday newspaper Direct mail In/on pack for own brand oucmwm #00900 H Source: A.C. Nielsen Company Price Prggotion and Brand Attitudes The majority of sales promotion today tends to be price oriented. For instance, couponing represented 703 (Figure 1.4) of all promotional spending in 1983 and 1984 and has steadily increased at the average rate of 16x annually over the past several years as Figure 1.5 shows ("Coupon Patterns," 1985, p.11). A recent study of consumers revealed that nearly 80% of the households use coupons when shopping 12 _ @ Refunds eo- \\ 65.72 \ 697° \69-675 Coupons N N N 20- \\\ . N 1 N I N 1981 1982 1983 1964 Fig. 1.4 Allocation of Dollars to Sales Promotion Techniques. Source: Dancer, Fitzgerald Sample, Inc., March 1985. (Nielsen, 1985). The effects of continued couponing, are unknown. As Quelch has stated, however, selling on the basis of price, as most promotions do, cannot have a positive effect on a brand’s image ("Coupon Patterns," 1985). The belief that frequent use of sales promotion "tarnishee a quality image" (Beam and Schaffer, 1981, p.49), represents a "quick fix maneuver that eventually leads to the loss of the basic values the companies try to build in their brands through advertising" (Giges, 1980), or weakens the brand differential which leads to a commodity—type situation (Christopher, 1971), has been behind the resistance to sales g 175 150 125 100 75 50 25 13 +14% NN NNN N 1980 1981 1982 1983 1984 Fig. 1.5 Growth of Couponing, 1980-1984 promotion. These comments from industry prove this belief it still prominent in the industry, and the fear is aggravated by evidence from case histories which seem to support this rumor. Just as advertising is thought to build brand images which add value to the product and create loyalty, sales promotion is suspected of damaging this image, diminishing the value and encouraging brand switching. Including sales promotion incentives, especially price promotions, in media advertising, subtly shifts the selling point away from brand attributes to the deal or discount. One British company adopted a policy of separating promotion 14 from regular advertising for just this reason. Their experience had been that the insertion of a promotional announcement in regular media advertising diminished the impact of those advertisements (Strang, 1975). Still, research on attitudinal effects of sales promotion is almost non-existent and remains inconclusive. Strang (1975) reports some evidence that shows increased use of sales promotion as opposed to advertising, damages brand attitudes over the long run. Tuck and Harvey (1972) find mixed results depending on product category, brand and promo— tional technique. Strang and Gardner (1983) find absolutely no effect on brand image in the short run. The frequency of sales promotions, not only across product categories and brands, but the repeated use for just one brand, is an important issue in this debate which unfortunately has not been addressed. Although advertisers have long been asking "how much is too much" regarding advertising frequency, the idea of wearout for promotion is new. The only published study on the attitudinal effects of advertising with coupons tested only one exposure to each advertisement. Effective frequency studies for advertising show the response to advertising doesn’t always accumulate positively over time. Similar response functions for promotion also need developing. One thing is for certain, even in spite of the lack of research, sales promotion has become a widespread and growing industry problem. As one executive predicts, breaking out of the ”discounting box" will be one of the most difficult 15 problems facing package goods manufacturers over the next decade (Meyers, 1985, p. 4). Study Purpose The purpose of this study is to gather some preliminary evidence on the effects of sales promotion on brand evalua— tions from a communication perspective by comparing media advertising without promotion to advertising which integrates promotion. This study will examine these promotional effects in a controlled frequency context for the first time. Organization of the Dissertation Chapter II develops a working operational definition for sales promotion and then systematically review the concepts basic to the design of this study. Several theoretical paradigms are applied to the sales promotion scenario and the implications discussed in terms of long-term brand image. The final section then outlines the selected research questions and hypotheses which are most appropriate for this study. Chapter III describes the research design, study procedures, and the sample in detail. The selection and characteristics of the test and control advertisements are also discussed. Independent and dependent variable manipula- tions, along with the coding schemes, are explained and the data analysis procedures are briefly summarized. Limitations of the study methodology are discussed. Chapter IV presents the empirical findings for each of the individual test product categories and discusses these 16 findings in light of market and consumer characteristics associated with each product category. Chapter V summarizes the findings. By comparing results across product categories for all dependent measures, some effort to draw generalizations is attempted and the importance of category differences will be explained. Directions for future research are also suggested. CHAPTER II REVIEW OF THE LITERATURE Sales promotion is a difficult concept to define because it describes a wide range of promotional techniques which exemplify diverse characteristics and which can be used to target very different audiences and address many different goals. The first part of this chapter compares several traditional definitions that have been used to distinguish between advertising and sales promotion. A simple, operational perspective of this sales promotion/advertising relationship is then provided which is useful for this exploratory study. This is followed by a review of key concepts important to this study. Specifically, a discussion of brand image, the main criterion variable under scrutiny, along with brand loyalty and deal proneness, two important covariates in the study of sales promotion or purchase behavior, is presented. All three variables are briefly defined, the basic literature relevant to these concepts reviewed, and the various means by which to measure these variables outlined in anticipation of the methodological considerations for this study. The bulk of this chapter, then, addresses the theoretical perspectives applicable to sales promotion and attempts to provide a framework by which the published 17 18 literature can be organized. This infrastructure will help to explain inconsistencies which have arisen in this research due to differences in the theoretical premises underlying each. Organizing the research this way will also make it more apparent as to how these theories relate to one another and how this exploratory study fits into the whole complex diorama. Finally, the chapter ends with a description of the research questions and specific hypotheses selected for this study. Defining Sales Promotion Product promotion in the generic sense, refers to the combined efforts of four promotional ingredients: personal selling, publicity, advertising and sales promotion. The research and literature described in this and the following chapters, address the two latter elements from this general promotional mix--advertising and sales promotion. An essen- tial first step in discussing the relative impact of these components is to establish operational definitions for each term or a paradigm which describes the underlying nature of each and what activities constitute one component versus the other. However, defining sales promotion as it differs from advertising is a difficult task. The American Marketing Association’s Committee on Definitions (1960) defines sales promotion as any activity which cannot be classified as advertising, publicity or Personal selling, as does a recently published Handbook of 19 Sales promotion (Ulanoff, 1985). Characterizing sales promo- tion as a catch—all for any activity which does not fall neatly into the definition for another category does not identify what sales promotion is, but rather what it is not, and provides little guidance in understanding the true nature of sales promotional techniques. Similarly, other authors (Strang, 1975; Block and Totten, forthcoming), position advertising as any paid form of non-personal communication under clear sponsorship which presents ideas, goods or services in measured media, such as television, radio, outdoor and print, and describe sales promotion as all other forms of sponsored communications including trade shows and exhibits, coupons, samples, premiums, contests, sweepstakes, cents—off packs, rebates, bonus packs, point—of—purchase materials, and trade allowances. This description, too, treats sales promotion as a leftover. What further complicates definitions like this is the implication that the mode of delivery is the differen— tiating factor between advertising and sales promotion. This definition suggests that advertising is placed in mass measured media and that sales promotion is unmeasured communications delivered by means other than mass media vehicles, such as in or on the package, in the store, or door—to—door. Determining what constitutes a mass medium is an initial difficulty here because direct mail can be classi- fied as both a mass medium and a type of sales promotion. In addition, as television, radio and print are increasingly used to support sales promotion events and below—the-line g; 20 price promotions, the mode of delivery is no longer a useful differentiation. Sales promotion has also been viewed as non—recurrent selling activity, ad hoc selling activity, or as any selling effort outside the ordinary routine. This definition is not as appropriate as it might have been once; for many brands today, sales promotion is an integral and frequent part of the promotional mix because marketers believe that to retain a competitive position against other promoted brands, they must "fight fire with fire." Sales promotion is perhaps most commonly described as a "direct inducement or incentive to the sales force, the distributor, or the consumers, with the primary objective of creating an immediate sale" (Schultz, 1982, p. 8) and as "short term incentives to encourage purchase or sales of a product or service” (Kotler, 1980). This is, of course, in contrast to advertising which is designed to create an image or personality for a brand over time. The factor differentia— ting sales promotion from advertising in this type of definition is the perceived promotional goal and time parameter involved. Advertisers use advertising for long- term attitudinal tasks and plan sales promotions for short- term behavioral needs. This perspective is narrow and, if taken literally, misleading; advertising, like sales promotion, can be designed to achieve short term objectives, as with direct response television commercials, while sales promotions can build brand image and provide excellent 21 long-term benefits. For example, premiums, carefully Selected, can carry out a theme and foster a desired image whet) linked with the brand in the consumer’s mind. Bantick (1980) calls for broader, more intelligent use of sales promotion, documenting promotions which have successfully achieved long term objectives. Yet, this short—term perspective is most prevalent. Although these definitions are not incorrect, and in composite, capture something of sales promotion’s nature, they paint a stereotypical picture which is not reflective of sales promotion’s complexity and harmony with the other asPects of the promotional mix. Advertising and sales promo- tion are intimately related, with many overlapping Characteristics; drawing concise boundaries between the two is difficult. Beam and Shaffer (1981) recognized this cCfilununality when they proposed a unique integrated marketing ElPpr‘oach interconnecting all marketing and promotional vari- 8131-ea. In their model, sales promotion is one of three modes in larketing. Figure 2.1 depicts these three modes and their relationships. The first mode is the product or service offer itself, taking into account the terms and conditions of the sale. The second mode represents all persuasive communications degigned and intended to enhance the basic offer by aiding the consumer to envision how he or she would benefit from its use. and therefore, stimulating desire for it. These include advertising messages, publicity, personal selling and a by nonverbal impressions communicated through promotional ‘, _ . 22 Iedia placement, retail distribution, or packaging. The thir‘d mode of marketing, promotional inducements, represents any activity designed to trigger action through a "carrot and stic:k" strategy. In other words, the consumer (or trade) recs: ives an extra something beyond the benefits of the basic offeerm—the carrot-—in return for certain concessions such as pur'c:hasing now instead of later, purchasing multiple units, or shopping a particular store--the stick. Me:§sages Designed to Extra Substantive Benefits Enliance Impressions 0f Basic Offer 0 Verbal O Nonverbal 0 Character of Basic Offer C'External to Basic Offer Basic Offer Regular or Standard Substantive Benefits .Ierms of Sale 0 Product Fig. 2.1 The Three Modes of Marketing ‘IIIIIIIIIIIIEZZ I . 23 In Figure 2.1, the three modes are portrayed as distinc- tly separate marketing activities with different functions but with overlapping corners. The model deviates from traditional attempts at distinguishing between promotional activities by acknowledging that sometimes an activity can be classified jointly. In this model, for example, packaging is part of the basic offer and distribution or delivery is a term of sale, yet each can persuasively contribute to the Product image and be considered persuasive communications as well; promotional inducements can become so tightly asso— Ciated with a product through continued use that they become an expected product attribute of the basic offer; media ac“vertising supporting sales promotional events and offers a10mg with a persuasive selling message contains elements of both persuasive communications and promotion. Beem and Schaffer’s paradigm is broader than necessary fol” distinguishing between advertising and sales promotion “ince it incorporates all marketing variables; yet, it over— comes the difficulties associated with black and white clansification schemes. Figure 2.2 diagrams the operational perspective used for 'I this study which is similar in conception to Dean and shaffer’s Third Mode of Marketing triangle. A simple pr“motion continuum is suggested with advertising and sales promotion anchoring the extremities as opposed to represen— tins two separate entities with definitive, mutually exclusive attributes. This conceptual continuum overcomes, in the same way the previous model did, definitional dilemmas 24 in instances where the boundaries between activities become °bscure, as with direct mail, or when the sales promotion Sales incentive is carried as a component in regular message advertising via a mass media environment. The continuum, however, visually extends the area jointly occupied by adver— tising and promotion which is probably more realistic. Co-parison among executions or tactical variations within either extreme, or among different combinations of adverti- . .. _‘/———\_ sing and promotion, is possible. In the Dean and Shaffer model, the legs of the triangle do not represent continuous (Padations of persuasive communications or promotional inducements. Instead each leg and corner constitutes a discrete category which fails to distinguish among such t1<>ns today. Sales Advertising Promotion tBetical variations and the relative nature of most promo- K Fig. 2.2 The Advertising-Sales Promotion Continuum f Using a continuum approach, a promotional piece is I defined in relative terms where the advertising extreme con— ‘titutes messages which persuasively convey an image about a product which personalizes the benefits derived (emotional or full(itional) and is usually delivered via the mass media. sales promotion represents a direct inducement offering an e. Rtra value or incentive to buy over and beyond the regular 25 Product, usually at the point of purchase. A print adverti— lemaent with a coupon in the corner is considered sales Prom-otion when compared to the same print advertisement witlmout any cents-off promotional deal; this ad with the COllI>0n would be considered advertising relative to a free standing insert (FSI) or direct mail coupon. Sales promotion can be targeted toward the consumers or the: trade and the two types of promotion have not been consi- det‘ted separately in the discussion this far. However, this research will concentrate on consumer promotions as they affiect consumer perception of brand image rather than on Sales promotions directed to the trade and distributors. Concepts Basic to the Study of Sales Promotion Brand Image Brand images represent the overall perception of a brand derived from existing information and past experience. Aa-ker and Myers (1975) capture the essence of brand image as follows: "every brand that has been on a market for any appreciable length of time begins to take on a personality or set of meanings by which it is known and through which people describe, remember, and relate to it....Such meanings are not inherent in the object itself but are ascribed to it by the people in a culture or society and by agents like advertising that can foster and support the impres— sions....An important human tendency is to reduce this vast array of meanings into some summary form or stereotype...." (pp. 139-140) lu‘ifi stereotype is a brand’s image. For products or services tllail are highly similar in functional respects, image can distinguish a brand from the competition by attributing 26 Perceptual or emotional differences to the brand. How a Consumer perceives a brand, in terms of or in spite of brand lttr‘ibutes, and the strength of those beliefs, depends on the abil_ity of the selected marketing and promotional techniques to i.mstill the image. Advertising is usually considered a powerful force in por‘tzraying brand image because it can directly show or huii.rectly imply the product in association with memorable sittmations, people, places, events, emotions, or outcomes, to estmablish imagery for a brand. Hence, brand image is asso— ciated with media advertising and not with asles promotion. David Ogilvy, a master of brand image in advertising, believes that all communication should contribute to the "Complex symbol" which is the brand image but thinks that adVertising, not sales promotion, does this best. "The manufacturer who dedicates his advertising to building the most sharply defined personality for his brand will get the largest share of the market at the highest profit....the manufacturers who will find themselves up the creek are those shortsighted opportunists who siphon off their advertising funds for promotions....A steady diet of price promotions lowers the esteem in which the consumer holds the product; can anything which is always sold at a discount be desirable? (Ogilvy, 1964, p. 412) This traditional attitude toward sales promotion under- lies the current concern about the effects repeated use of prOlllotions may have on brand image. Prentice (1975) challenges this traditional view which ‘nitfi advertising against sales promotion with respect to b r‘ind image. All promotions are not price or discount 0 r1ented as he points out. He instead recommends a consumer 27 franchise building (CFB) approach which hypothesizes that "to be profitable in the long run, a brand must build and maintain a significant, lasting value (emphasis added) in the minds of...consumers" (p77). This lasting value is equated to the brand image and any activity which helps implant such long term impressions is a consumer franchise building effort, including advertising and some types of promotion. In fact, Prentice has divided sales promotion techniques into those which are consumer franchise building and non—CFB. Price promotions, such as coupons, not accompanied by a Iessage about the brand or its attributes are not considered Consumer franchise building. _The effects of sales promotion on brand image The published literature concerning sales promotion effects on brand image is almost non-existent. Strang (1975) reports a case history of two British toilet tissue brands, Al'lClr'ex and Delsey, from 1958 to 1969. Andrex, with slightly hiSher market share than Delsey, continued to gain market Share over the twelve year period by maintaining its regular adVertising in the face of intense competition. Delsey reSponded to the competition by adopting a sales promotional “trategy and cutting advertising. This resulted in an i"mediate and steady decline in market share. More to the point, longitudinal attitude research on these brands showed that between 1964 and 1969, the percentage of housewives who evaluated Delsey as soft and atl-"ong, decreased from 33 to 20 percent, while during the 28 Same period these characteristics were increasingly attributed to Andrex, by 40% increasing to 55%. Such a finding implies a direct weakening of brand image or loss of has :ic value along key attributes to the brand which promotes at the expense of advertising. The toilet tissue example ties the declines in market performance to attitudinal problems which the other sales— oriented case histories reviewed earlier can not do. However, this study is plagued with the same complications and uncertainties associated with aggregate expenditure leasures and lack of control over extraneous variables. Stuclies designed to control for outside influencing factors are more appropriate to explore the validity of claims such as those this study makes. Hamm, Perry and Wynn (1969) conducted a controlled SC‘lomon Four Group experiment with a random sample of almost 3500 male students to investigate the effects of sampling on illage and purchase motivation. Men’s hairspray samples were distributed among the respondents and a pre-post comparison sh<>aned significantly positive results on both the image and l"-llr‘chase intent. Tuck and Harvey (1972) showed money-off and premium promotions to have a direct negative affect on brand attritudes for two brands of washing powders, and a positive effect on milk food drinks such as Ovaltine. They concluded tha1: it is difficult to predict sales promotion effects since these effects vary by brand. Furthermore, different premium Offers in this study fared differently and inconsistently 29 with expectations. This underscores the necessity of praztesting and cautions against generalizing results of one pr<>motion to all sales promotion. The recent Stout and Leckenby review of advertising copy reesearch (1984) revealed no studies which addressed price or pr‘i.ce promotions as selling messages in advertising and the ef’iiect of this on brand image or other advertising effective- ness measures. One study not included in that review, hOWever, did address coupon effects in advertising. Strang and Gardner (1983) looked specifically at identical print ads With and without coupons and measured the effect on brand ratings for (l) Malt—o—Meal cereal, (2) Style Shampoo, (3) I-il>ton Herbal Tea, and (4) Tostitos Tortilla Chips. He found no consistent or significant results to suggest that COUponing damages a brand’s image. Strang and Gardner (1983), however, did not test couponing over repeated expo— Bl-nt‘es which has been found in advertising studies to alter effectiveness levels . Measurin brand ima e Since the decision to develop a brand’s image is likely °°upled with a desire to strategically differentiate the ,1 brand from competitors, and since that image is a composite ‘yibol based upon the perceptions of its combined characteri- atice, measurement of brand image has taken two parallel r‘omtes. The first is a comparative approach while the second all’Droach is non-comparative. 30 The comparative methods involve assessing the way one brand is positioned or perceived by consumers versus the competitive brands, usually on several important dimensions. Multidimensional scaling (MDS) is an example of the compare— ti ve approach. MDS employs a paired comparison technique to evaluate the similarity of all brands to each other and, sometimes, to selected attributes of importance. The resul- ting perceptual map plots out spatially the relative percep- tual differences among brands. Smith and Lusch (1976) simply had consumers group brands according to similarity. Because consumers rarely make a purchase decision in isolation from the competitive alternatives, and because the purchase deci sion is a multi-faceted process, this method is real istic. The primary problem is in selecting the apl-“'l“0priate brand set for evaluation. With the non-comparative approach, consumers evaluate a bx‘tElnd along several attributes, usually with conventional Bcal ing techniques. A difficulty with this approach is in sele(:ting the most appropriate attribute(s) to scale. Aaker and Myers (1975) warn that the adjectives should be e,':l“&ustive, meaningful, interpretable, and relatively 1“dependent. Strang and Gardner (1983) assessed image in this way. For each brand in question, a different set of semantic differential scales were administered and the individual attribute ratings compared across promotion and non-promotion <2 Onditions. Tuck and Harvey (1972) measured image effects V 1‘ an overall "Good/Bad" ll-point scale, and then combined 31 this with a series of ll—point bipolar adjective scales which scaled evaluative beliefs about specific product attributes. In another study, subjects were asked to scale their agreement or disagreement to a series of evaluative state- ments about the brand in Likert scale fashion (Perry, Izraeli, and Perry, 1976). Hamm, Perry and Wyn (1969) looked at the effects of a free sample of hairspray, a product perceived to be feminine at the time, on males’ image percep— tions. Image in this case was measured via a nominal categorization of the product as masculine, feminine or neutral. In either the comparative or non—comparative approach, leasurement of brand image is usually tailored to fit the sPeeifications of the study hypotheses or the characteristics °f the product being studied. Therefore, it is often difficult to compare across studies or product categories where different adjective scales, brand pools, or Likert Statements were used to measure the image. Brand Loyalty Brand loyalty denotes a bias toward or a commitment to Q pQrticular brand by a consumer over time, usually ‘Qnifested in the consistent repeat purchase of that brand. In the absence of brand loyalty, it is assumed that a § Oheumer would merely distribute his or her purchases of a b b0duct among all the various brands available in a purely b andom manner. 0n inspection of a consumer’s purchasing gcord, then, finding any regular pattern where any one brand 32 has been consistently chosen over another would indicate that some kind of rational, purposeful selection is occurring, signifying a degree of brand loyalty. This approach to brand loyalty is purely behavioral. Yet repeat purchasing behavior can occur out of loyalty toward any number of factors. For instance, consumers may be part ial to a certain retail store which only stocks one brand, effectively excluding other competitive alternatives from the choice consideration. Loyalty toward a particular PFOduct attribute such as low price or package design might dictate the continued purchase of one brand, but only so long 88 it is the most economical or the only alternative in the desired package. Where many brands exist, consumers might rePurchase one brand to simplify the search task and reduce the risk involved in the selection process. In fact, R°3€=1ius tested eleven risk reduction methods and found brand 1"Yalty to be one of the most favorable and frequently used atrategies for risk reduction by consumers (Roselius, 1971). While the behavioral outcome of these factors is repur- chase of a brand over time, should circumstances change——for jinstance, a new brand is introduced or a competitive brand tQ‘Porarily price discounted--a consumer would likely shift 1 oyalty to another brand. Nevertheless, behaviorists would argue that ”brand <>yalty is always a biased response to some combination of §haI‘acteristics, not all of which are critical stimuli" ( TQCker, 1964, p. 32), but that these issues are not of § QhBequence . k 33 ”No consideration should be given to what the subject thinks or what goes on in his central nervous system; his behavior is the full statement of what brand loyalty is" (p. 32). Jacoby (1971) feels that a psychological commitment to a t>r‘and, in addition to a behavioral response, is necessary to (iistinguish between repeat purchasing and brand loyalty. ”To exhibit brand loyalty implies repeat purchasing behavior based on cognitive, affective evaluative and predispositional factors--the classic primary components of an attitude" (p. 26). Day (1969) similarly suggests that there is a critical difference between "intentional” loyalty and "spurious" 1037£31ty, a label he gives habitual purchase of a brand absent Of strong feelings of commitment toward the brand. Day’s tw<>-dimensional approach to brand loyalty involves both brand 1°Ya1 behavior and brand loyal attitudes, or to a deliberate decision to repurchase a particular brand because of some ree11_ or imaginary superiority attributed to that brand. Although social psychologists such as Olson and Jacoby (1971), Day (1969), and Jarvis (1978) have examined the brand IOYSlty concept from a psychological perspective, the .arketing literature has concentrated on the behavioral asPect of loyalty. Given the nature of the panel data that 11“ been used in past marketing research regarding consumer ]L<)3’alty, distinguishing among psychological commitment and ‘:‘1€3 different behavioral loyalty types has not been possible ‘itld has, as a practical matter, not been the key concern. it? 1‘63 published literature on brand loyalty has focused more on ‘dl‘3<=umenting the existence of the loyalty phenomenon and La 34 evaluating its usefulness in marketing segmentation than on the motivations behind the loyal behavior. Brand loyalty correlates From the marketing literature, it is evident that —‘W" lc>3r£alty to a brand within a product category does exist. Senreeral studies of 1950 and 1960 panel data available from the: Chicago Tribune, J. Walter Thompson and Market Research Corporation of America (MRCA) data bases, investigating nulleerous product categories, have shown that only a minority 0f ti category’s purchasers are 32; brand loyal to some degree, while the majority concentrate purchasing among one 0? trwo favorite brands (Brown, 1952; Cunningham, 1956; Guest, 1955). Cunningham (1956), for instance, found that 9036 of 'Cnnte families’ purchases over three years were for one brand in E1 category. This tendency to favor one brand was found to °c<>UJ‘even with fictitious experimental brands of identical quality and appearance. Over half of the subjects in the e"‘Periment developed a degree of loyalty greater than could be expected by chance, suggesting that, for some consumers, loyalty is just human nature (Tucker, 1964). However, brand 1olfal proneness, when a family is brand loyal for all I>ut~chasing, is rare (Cunningham, 1956). Furthermore, in most I>t“<>duct categories, a brand’s sales seem to be spread evenly “Ong consumers with all degrees of loyalty rather than QOhcentrated among either extremely loyal customers or with ‘Witchers (Cunningham, 1956) . 35 Although many researchers have tried to link brand loyalty to socioeconomic, demographic and personality variables, these inputs have had virtually no explanatory power. For example, a 1964 study by the Advertising Research Foundation of toilet paper loyalty found only .053 of the variance in purchasing behavior could be explained by 15 socioeconomic variables plus personality variables (ARF, 1964). Massey, Frank and Lodahl (1968) found 103 of beer Purchasing could be explained by a combination of socioeco- no-ic and personality variables but less than 53 of the variance for coffee and tea could be predicted. Farley (1963 and 1964) could explain only .043 over numerous grocery Products. Kuehn found a possible explanatory link with purchase aCtiVity. For orange juice, it appeared that heavier Purchasers demonstrated higher loyalty than light users (Frank, 1967). However, the findings of Cunningham (1956) find Massy, Frank and Lodahl (1968) are inconsistent with this. This discrepancy could be contributed to measurement differences or perhaps product category differences. Frank exPlains Kuehn’s results by the fact that frozen orange juice was relatively new at the time of the study, suggesting that this correlation would not be apparent with established brands. Given the evidence, it would appear that purchase QQtivity, too, provides little predictability of loyalty and fit best, may be slightly correlated in some product Categories. 36 German (1970) has suggested that a consumer who restricts the number of stores visited, thereby restricts the Opportunity to be disloyal. Under the assumption that "store loyalty is a regulator of brand loyalty," Carman hypothe- sized that the single most important predictor of brand loyalty would be store loyalty. He found that store loyalty exp lained over 63* of the variance for coffee, canned fruit, and frozen juice. Cunningham (1961) looked at this same relationship and, although he found some explanatory power for three out of 18 products when testing on a product—by- PPOduct basis, he concluded that for most purchasing, store IOYalty does not go hand in hand with brand loyalty. A last Study comparing those who purchase private label goods and thoese who primarily purchase branded products, did find that loyal private label users tended to shop stores which carried Private label products more than other customers (Frank and Boyd, 1965). Cunningham (1956) showed dealing to correlate primarily with low brand loyalty and brand switching. This finding is <2‘t’ueistent with several other studies which also found that Clealing is not as influential with brand loyal persons as it 13 with brand switchers (Tucker, 1964; Frank, 1967; Frank, Massy and Morrison, 1964; Massy and Frank, 1965). %uring brand loyalty All the research has shown brand loyalty to exist; ()‘Vever, the translation of the conceptual definition into an Q berational measurement tool has been inconsistent. 37 Measuring brand loyalty based on the sequence of brand cb<>ices over time is one of the first measurement schemes useeci. One of the earliest studies, Brown’s analysis of the Ciliszago Tribune data in a sequence of Ad Age articles in 1952 atici 1953, measured brand loyalty by categorizing the different types of purchasing patterns exhibited by family diary data. Those families with undivided loyalty purchased tine: same brand repeatedly in succession over the study period wtli.]_e divided loyalists alternated purchasing between two fenxr<>rite brands in a regular way. Families exhibiting brand restlessness could be described as having unstable loyalties; these families would be completely loyal to one brand for a BIIC>z-t period but then would switch to another brand for BWIli.le, and then again to another. Anyone not falling into °n<3 of these loyalty categories was classified unloyal and uslltllly had a more random purchasing pattern. Below in Table 2‘ 1 , Brown’s categories are diagrammed for better understan— ‘iitlg. Tucker (1964) also looked at the sequencing and frequency of purchase over time. He classified families as b1"and loyal if they recorded purchase of the same brand at least three times in succession. TABLE 2.1 BRAND LOYALTY MEASURES IN BROWN’S STUDY Sequence of Brands Undivided Loyalty A A A A A A Divided Loyalty A B A B A B Unstable Loyalty A A A B B B No Loyalty A B C D E F g 38 The extent to which purchasing was concentrated on a few brands in a category is the basis for the market share brand loyalty concept used in numerous studies (Cunningham, 1956; Massy and Frank, 1965; Frank and Messy, 1965; Frank, Massy and Morrison, 1964). The proportion of total purchases represented by the largest single brand used of all brands bought over the period under study was used to figure brand market share. Cunningham (1956) and Massey and Frank (1965) also percentaged the total purchases for the two most- purchased brands to calculate dual loyalty market share. Farley (1964) simply counted the number of different brands bought in the period, as did Massy, Frank and Lodahl (1968). A third technique for measuring brand loyalty is to look at the stability of the favorite brand over time. Usually in t"his case, the data is split into two halves and the most “38d or favorite brand of each period compared. If the brand ”33 the same across the two periods, the family was consi- dered brand loyal (Guest, 1955; Massy, Frank and Lodahl, 1988; Farley, 1963—64). These studies of the 1950’s and 1960’s reviewed here and by Frank (1967) were conducted with data from consumer diary I’al'l-els and all have a similar analytical problem: purchasing §et: Percentage of purchases devoted to most frequently purchased brand and to top two brands Cunningham (1956) Frank, Massy & Morrison (1964) Percentage of purchases devoted to most frequently purchased brand Frank & Boyd Purchased test brand more than any other Frank & Massy (1965) Massy & Frank (1965) Diversity of brEU31ex trade-offs involved in the selection process. Short of Changing or revoking the decision, this need to reduce the lashitude of the dissonance will lead the individual either to Qhangs the cognition about the chosen and rejected brands “nt 11 consonance exists, or to establish cognitive overlap (g ihilarity) between the alternatives. In other words, an 1 IQ W! 4..) -H 8 Positive Learning m Factor ‘\ \\ Net Effect \ g Repetition \\ Neutral \ Factor Negative Q_ Fig. 2.5 Two-Factor Theory of Affect Repeated Stimulus of the learning curve. Negative Tedium Toward a The response decay these of these variables closely followed and paralleled the decay in attention. When investigating learning in same commercials, Grass and Wallace (1972) results were mirrored in the real world. In a well documented lab study by Ray creative execution for two product ads was one ad was attention-grabbing and the other was not. the field on these found these and Sawyer (1971) varied such that The grabber ad scored much better than the non~grabber ads in recall, which in this one case implied that the more the 70 attention paid to the ad, the greater the learning of its content. Attention in this study was linked to execution manipulations. The attention value of an advertisement is certainly influenced by an ad’s creative qualities; therefore, creative approaches which enhance attention may increase learning and brand attitudes. These execution differences were ignored by Grass (1968; Grass and Wallace, 1972). Ray and Sawyer (1971) did not elaborate as to what differentiated the grabber from the non-grabber ads. Silk and Vavra (1984) compared hard sell to soft sell advertising approaches, predicting greater attention levels and corresponding recall levels for the hard sell than for the soft sell. Their hypothesis was partially confirmed. After a single exposure, recall, awareness, and brand preferences were significantly greater for the hard sell ad. After two exposures, although the hard sell condition still led, the margin was significantly reduced because the incremental increase was greater for the soft sell than for the hard sell advertisement. Ray and Sawyer’s findings with respect to grabber and non-grabber eds showed similar patterns over repetition. In these studies, increases in attention corresponded to increases in learning and attitudes, while declines in attention were accompanied by decreases in these measures. Grass hypothesized that the key to maintaining attention levels and preventing wearout was to build a campaign around several alternate executions to limit the tedium to any one 71 execution and keep attention levels high. One particular field experiment showed that, as predicted, substitution of different ads regenerated interest or attention and slowed decay of recall by more than 50X (Grass, 1968). Ernst (1978) has also shown that varied advertising messages for a brand elicit greater recall over repetition and maintain response levels over a greater number of exposures. Greenberg and Sutton (1973), after reviewing the advertising wearout research, recommended adding several claims or nuances to one advertisement to lengthen the learning process and prolong the time period before any negative affect would begin to accrue over repeated exposure. Although inattention is only one plausible factor to explain learning decay, this along with the two factor theory can be applied to any repeated stimulus. Unless coupons within an advertising environment affect the attention level, the length of the learning process or the creative tedium factor, the results for this type of sales promotion would be no different than for regular advertising. However, based on Starch and Gallup and Robinson readership scores, Bowman (1985) has reported that coupons increase the level of attention given to an advertisement. At the very least, coupons might represent an attention- grabber to some people--those who use them, for instance. Hence, for deal-prone persons, coupons may attract attention, sustain attention longer, prolong learning, and reduce tedium. If Bowman is correct, coupons may also serve to gain the attention of non—deal prone persons. However, a coupon 72 would not be of much interest to someone who does not deal and the ad may lose attention quickly after the initial exposure. In fact, a coupon may add to the tedium of an advertisement more quickly, shortchanging the learning process and resulting in satiation sooner than otherwise. After all, Silk and Vavra (1984) found negative effects for the hard sell advertisement after the initial exposure as compared to the soft sell approach. Likewise, Ray and Sawyer (1971) saw reduced effectiveness for attention—grabbing ads over repetition as compared to non—grabber, more subtle ads. If coupons can be described as hard sell or attention- grabbing, these studies make a clear statement of the impact on a brand image couponing would have over repetition-- diminishing returns approaching negative response. Multi-attribute models of attitude Since learning is a continuous process needing repeated reinforcement, and attitudes are learned predispositions, it should be possible to alter attitudes through exposure to the ”learning experience" of a persuasive message. Multi- attribute models take such a perspective. These paradigms quantify consumer attitudes by breaking them down into a summative model of cognitions or beliefs about a brand with respect to its attributes, each attribute weighted according to the salience it has for the individual. According to a multi-attribute model, to change an attitude an advertiser could either shift the importance of an attribute or change the beliefs about how the brand rates on that attribute. 73 Fishbein and Ajzen (1975) have proposed a multi- attribute model in which price is hypothesized to always have an effect on attitude formation. A brand’s ”value” is deter- mined as a ratio of a brand’s perceived attributes weighted by the salience of that attribute to the consumer, divided by a brand’s price weighted by the salience of price. If price is affected through a discount, this would affect a brand’s value. When perceived price is smaller, the calculated value, everything else constant, would increase. 0n the other hand, if the salience of price is increased through continued promotion, this would offset the added value of a discounted price and, in changing the ratio, decrease overall brand value and the relative contribution of other attribute evaluations (Sawyer and Dickson, 1984). Information Processing Consumer information processing refers to the way consumers attend to, search for, organize, store and retrieve information from the world around them. It is not so much a theory as it is a perspective. Consumers are viewed as processors of information with internal and external memory stores and limited processing capabilities. This leads to selectivity in what they perceive, how they interpret infor— mation, and what and how that data is stored. In many respects, information processing is similar to cognitive learning theory. It simply represents a process-orientation based on how learning is accomplished given the way the human 74 memory processes, stores, and uses information from a message or an experience. The focal point of the information processing system is the memory. Obviously, memory plays a major role in consumer decision making and Bettman (1979) has written of memory factors involved in consumer choice strategies. He reviews three different, but not necessarily incompatible, descriptions of the structure and role of memory in consumer choice situations when information is processed. The multiple store model proposes that everyone has two separate memory stores with distinctly separate functions. The short term memory store (STS) has only limited processing capability and sensory input data first enters here. Proces- sing capacity limits the amount of information that can be attended to at any one time but information that is processed can be transferred to a second, permanent informa— tion repository--the long term store (LTS). This store has unlimited storage capacity and information can be retrieved at any time and brought into STS to aid in interpreting new stimuli or in making a decision. A second model-~the activation model--assumes only one memory store, of which only limited portions can be activated at any time. Activation is usually specified in terms of rate or intensity, which determines not only what is stored, but how the information is stored. The third paradigm also suggests only one memory store with limited processing capacity. Craik and Lockhart (1972) have proposed, however, that this capacity can process at 75 different levels. Lower levels of processing would be limited to very superficial sensory data absent of any inter- pretation or complex analysis. Higher levels of processing would involve more complex semantic and cognitive elabora- tions on the information to integrate the new data with existing beliefs and knowledge. The higher or deeper the processing, the more details that could be retained and the longer the retention of that material. Whether or not the memory truly consists of two physio- logically separate structures or is one store in which STS represents a temporary activation of the information stored in LTS, the three models agree that retained data in a memory store consists of semantic concepts or experiences with associations among them. Events and episodes can be stored in a time and place oriented context called episodic memory (Tulving, 1972), or this information can be organized in a meaningful way and integrated with stored data on past experiences and expectations. This internal structure, the memory schemata, is usually described as a network of nodes and associative links among these nodal concepts. It is the strategies used to manage the flow of informa— tion in and out of the memory that are most important. These control processes concern how information is screened and structured or encoded for processing; how once it is structured, it is processed or ”rehearsed” for long-term retention; how it is transferred to permanent memory storage and in what form it is stored; where it is placed within that storage capacity in relation to other information; how acces- 76 sible it is for later retrieval in similar or different contexts; and how the information is reconstructed at a later point for use. Screening and encoding sensory input Upon exposure to an advertisement, an individual will attempt to process and retain the information presented. Where there is a lot of sensory input, limited processing capacity only allows some of the information to be encoded. Exactly what is encoded or selectively perceived versus selectively screened out will depend on how relevant the data is and how well it concurs with existing information already stored in memory. Katz (1968), for instance, showed that consumers seek communications which support their beliefs so as to maintain cognitive consistency. Brand loyalty might affect the way an individual screens the data in an advertisement for rehearsal. A consumer who is loyal to Brand X may screen out or give little energy to processing information on Brand Y received through adverti- sing, and have increased retention of messages related to Brand X. Deal proneness may also influence what is attended to. While Bowman (1985) reported that, in general, ads with coupons generate more ad readership, this may be mediated by propensity to use the coupons. Deal prone persons may pay more attention to advertisements with coupons than ads without, and be more likely to remember a coupon in an ad. 77 Strang and Gardner (1983) report, too, that elements within an advertisement can attract attention away from other features in the ad. If coupons draw attention away from other message elements, learning of relevant brand information and value perceptions will be reduced greatly. How much of the information is coded for rehearsal and transfer to long term memory storage is dependent to a degree on how it is presented and in what contexts. An advertise- ment which organizes the information effectively or presents a persuasive argument in a logical fashion will speed and facilitate the encoding process for transfer into LTS. If too much information is presented, the overload will result in only partial encoding. Studies have found that although consumers may be more satisfied with more information, they actually make poorer decisions with more information (Jacoby, Speller, and Kohn, 1974; Jacoby, et al., 1974). To reduce overload, and facilitate encoding, rhyming slogans or jingles which place an external structure on the information have been found to be beneficial (Howard, 1977). Imagery, too, has been shown to be an effective coding facilitator because an image can chunk many separate pieces of information (Lutz and Lutz, 1977). Whether a coupon in an advertisement hinders or aids in the encoding process for a brand is unknown, but as it rarely is integral to an ad’s message, a logical deduction would be that a coupon adds to clutter, increases overload and reduces processing of other ad information, such as intrinsic brand attributes. 78 Rehearsal, storage, and retrieval Rehearsal can be either pure silent repetition of the data (rote memory) or a more detailed mental analysis of the information and its consistency and/or relation to already existing memory schemata. What an individual expects to do with the information in the future will influence exactly what is processed and transferred to memory. For instance, a purchase decision made in the store, with brands as cues and the packaging or labeling serving as external memory stores for detailed information such as nutrition and caloric content, requires less rehearsal of such information and limited storage; a consumer need only to recognize the brand. If, on the other hand, a decision to purchase is usually formulated outside the store, the consumer will need to transfer more details into internal memory store to facilitate future recall and evaluation. This will require more information processing. The rehearsal and retrieval aspects of information processing were at the core of a study by Saegert and Young (1982) which tested the recall and recognition of print advertisements from a "levels of processing" memory perspective. After exposure to each ad, experimental subjects were asked a question designed to affect the type of rehearsal and to act as a memory cue for that advertisement. Questions which treated the advertised brand in a non-semantic context, or which indirectly related to the brand or brand attributes, represented cues thought to initiate shallow rehearsing and processing. For example, shallow processing was spurred by 79 the question "Is the brand name in blue letters?" To initiate a detailed rehearsal and deep processing, questions such as "Have you heard of this brand before?" which related the brand to existing memory schemata, were designed to require deeper analytical processing. Subjects in the deep processing condition recognized and recalled the brand significantly better after both one and two exposures. Storage of the data in the LTS is conceptualized as a network of nodal concepts linked together through learned associations with other nodes. Where and in what context(s) new data is placed or associated will influence recall. The strength and number of the links between new information and existing schemata will determine the ease of search and retrieval of the information later. Forgetting is viewed from the information processing perspective as a failure in retrieval rather than decay or extinction of knowledge. For this reason, aided recall tasks are thought to be superior to unaided tasks in retrieval because aided cues simplify and guide the search within the memory schemata to a greater degree (Tulving and Pearlstone, 1966). Information which is linked solidly with relevant concepts through the learning process, or which is integrated through associations formed within different contexts, will be retrieved more quickly than concepts which have been only vaguely learned in one, single cue context. Associations are strengthened through frequent reinforcement; therefore, ideas which are not often used or reinforced will become more difficult to retrieve through time as the bonds weaken. -*—_——____—‘ 80 Rokeach’s concept of belief centrality meshes with an information processing perspective on memory structure. Rokeach defines attitudes as ”a set of interrelated predispo- sitions to action organized around an object or situation" (Rokeach, 1968, p. 12). In essence, Rokeach proposes that our attitudes are linked together in a complex interconnected network of belief statements. In this network, some beliefs are central to our very being: beliefs about who we are, why we are, about the world around us. These core beliefs are easily defensible and very difficult, if not impossible, to alter or change. Other beliefs are peripheral to our existence, such as evaluations on brands and products we buy. Peripheral beliefs are not as defensible and are vulnerable to change given persuasive messages. To the extent that peripheral beliefs can be linked to innermost core beliefs or needs, then the less subject they are to change. Brand image is the advertiser’s attempt to link the product and its attributes to basic needs and wants with which consumers identify and which are central to a consumer’s belief system. Grass and Wallace (1972) found that attitudes decay at a slower rate than recall when inattention prevents reinforcement of the message. This finding, too, supports this memory structure. Attitudes would be perceived as being integrated more thoroughly in memory because the evaluation process whereby one arrives at a positive or negative predisposition requires extra processing of the facts in terms of more basic needs. 81 In cases of incomplete storage or retrieval of informa- tion, individuals will attempt to reconstruct the memories based partly on what was and an assumption of what must have been. This process of "response generation" coincides with the attribution theory. Several implications arise when relating this discussion back to sales promotion. First, if a coupon in an ad initiates shallow processing, brand name information may not get integrated into existing schemata well enough for later retrieval. Second, linkages which are reinforced frequently will prevail over those which are not. Therefore, brand associations which are strengthened at the expense of basic salient attributes will decay brand value in the long run. If messages about a brand are primarily price oriented, and the discount in the advertising focuses attention away from, or reduces processing of, brand imagery or intrinsic attributes, brand image may be weakened. Linkages between brand name attribute evaluations and core beliefs which are not established, cannot be retrieved. Information which is not reinforced will fade and deter access to this information. Third, if memory is truly structured as a network, then one concept may be indirectly linked to a great many other nodal concepts which have linkages in common with it, i.e., particular brand attributes or similar dealing patterns. An ad with a promotional element will link the brand name with the deal in addition to, or in place of, its attributes in 82 the LTS. Should a dealing technique be associated in that person’s mind with negative concepts such as "gimmick," ”cheap," or "low quality merchandise” from another experience, when it comes to retrieving information on that brand, the stronger links in the network will lead directly to the dealing aspect and of the brand indirectly to the retrieval of cheap, gimmick, or low quality. Cognitive responses Cognitive responses are the result of the rehearsal and structuring activity described as information processing. These subvocal responses to messages represent another difference between learning theory and information processing theory, although it is really only a difference in the measurement process. One consistent pattern among hierarchial learning paradigms is that the measured dependent variables tend to be intermediate response outcomes such as attention, recall, knowledge, attitudes, purchase intent, or sales, all which denote message retention or yielding. These hierarchies have been critically characterized as "black box" approaches (Wright, 1973). Information processing, or cogni— tive response theory as it is sometimes called, focuses on the processing engaged in by individuals upon exposure to a message rather than on just a measurable end product (Belch, 1980). Macoby (1965) and Greenwald (1968) recognized the promise of this approach in communication studies when they proposed that any response generated by a receiver 83 during the rehearsal and reception of a message, beyond the message itself, impacts on the final response outcome. These cognitive responses include recognition, associations, elaborations, ideas and images resulting from exposure to a message (Cacioppo et al., 1979). Such spontaneous thoughts have been found to mediate the effects of the message. In other words, the thought about the message when attempting to relate the new stimulus to the old belief structure, as well as the message itself, influences the message stored in memory. Cognitive responses have been categorized in many ways with many different labels and points of overlap. However, four basic classification schemes can be identified: (1) whether a response is positive or negative, (2) whether the response is stimulated primarily by the message or by the individual, (3) whether the comment is directed toward the content of the message or the source of the message, and (4) whether or not the thought is relevant or irrelevant. The most prevalent categorization scheme focuses on two types of valence comments: counterarguments and support arguments. A counterargument is produced when the incoming information from a message is discrepant with an individual’s belief system or with prior learning. This critical thought is assumed to counter the message claim and reduce the persuasiveness of the message. Support arguments represent "congruent associations" with already entrenched beliefs (Wright, 1973). Cacioppo et al. (1979) have noted that cognitive response research has consistently revealed a 84 positive relationship between the valence of the argument in response to a message and the yielding to that message. Cognitive responses have also been characterized according to the origin of the response. In other words, thoughts which are stimulated directly by a message and which comment primarily on information contained within that message, are message-related thoughts. Thoughts that only indirectly stem from the message content but which mainly involve inferences or associations generated by the ”own" thoughts. individual, are "recipient-generated" or The research on this has produced conflicting evidence as to which of these actually is more important in mediating message acceptance (Belch, 1980). Categorizations with respect to the target of the response have been inconsistently defined. Much of the research to date using this thought monitoring approach have been in the communications field and the persuasive communication has involved an issue such as anti-smoking, tuition increases, news censorship, the legal driving age, a ban on alcohol, or specialized education. In these issue- oriented studies, these responses have usually been divided into those which are related to the arguments surrounding the issues and those which are directed to the source of the message, usually in terms of spokesperson credibility. Because so little cognitive response research has been conducted in the advertising area where the message is an advertisement, coding the target aspect of the message is relatively new. For a product message, thoughts can be 85 directed toward the message or product, the advertiser, or the advertisement. Belch (1980) is one of the first to include ad evaluations as a separate cognitive response indicator. Defining what is relevant and what is irrelevant has been subject to each researcher’s own particular frame of reference. Irrelevant responses have included those which are ad-related, source-related, repetition related, and those which are recipient-generated ("own" thoughts), to name a few. However, Shimp (1981) has concluded that ad evaluations are not irrelevant in a consumer’s brand choice behavior, and Rethans, et al. (1986) and Belch (1980) have found evidence to support this. Repetition and cognitive response Cognitive response has been shown to vary with repeti- tion. Calder and Sternthal (1980), for instance, found that over repetition to the same message, the number of negative thoughts increased significantly and the yielding to that message decreased. Cacioppo and Petty (1979) examined messages that were either consistent or contrary to an individual’s beliefs and found that positive support arguments increased initially and then decreased with repetition for everyone regardless of orientation on the issue of a state tax. They, too, found that a positive relationship between the valence of the generated response and the agreement with the communication. In other words, message agreement first increased when 86 support argumentation was highest and then decreased when counterargumentation replaced the supportive response. Cacioppo and Petty (1980) also found that over four exposure levels (1, 3, 5 and 10) repetition led to increasing, then decreasing acceptance of two separate issues (an alcohol ban and the legal driving age). This was paralleled by increasing and then decreasing favorable thoughts (support arguments) and, obviously, initially decreasing and then increasing counterargumentation. Rethans, et al. (1986) investigated the response to television commercials over repetition. They applied cognitive response to the two-factor theorem presented earlier. They measured both ad—related and product-related cognitive response valences over repetition of a television commercial for a new product. They found product-related support and counterargumentation to have no effects on brand attitudes, or purchase intent over repetition. Ad-related cognitive responses increased significantly over exposure levels, indicating a positive learning factor does exist. However, this learning did not in turn increase brand attitudes or purchase intent. At the highest exposure level, negative repetition—related thoughts increased which shows evidence of the tedium factor described in the two-factor theory. Calder and Sternthal (1980) also conducted an experiment on the role of cognitive response in message wearout, only they looked at the role of message—related versus own thoughts in this phenomenon. They found that the number of own thoughts 87 increased and message-related thoughts decreased over repetition. Where repetition was higher and own thoughts were dominant, brand evaluations became increasingly more negative. These patterns persisted even when attention was sustained over exposure levels. They concluded that wearout could be the result of this shift from cognitive elaboration on the message and brand to elaborations unrelated to the message. As these selected studies show, considerable evidence exists which verifies that a cognitive response approach offers additional insight into the communication learning process. Since this measurement technique has not been utilized much in advertising research, it is still too early to know how much influence these processes have on brand choice situations. However, if a coupon generates either more or less favorable cognitive response than regular advertising, this might affect yielding to the message or liking of the advertised brand. Distraction theory Distraction theory, predicated on a cognitive response perspective, has been hypothesized by several authors to explain weakened resistance to advertising or promotion. Wright (1973) found that evaluative cognitive responses such as supportive or counterarguments generated by an individual in response to an advertising message, mediate the message persuasiveness, as well as the brand liking and purchase intent. To the extent that counterarguing or supportive 88 response is disrupted when attending to an advertisement, the message will become either more persuasive or less persuasive, respectively. It has been shown that as the distraction levels are increased, the number of counterarguments or cognitive response decreases, and yielding to the message increases. (Petty, Wells and Brock, 1976). Petty, Wells and Brock (1976) found that a logically sound, easily defensible and compelling message resulted in less counterarguing and more supportive arguments than a less compelling, easily refuted message. Distraction from these messages, then, resulted in the logical message becoming less persuasive and the less compelling message becoming more persuasive. If a coupon in an advertisement represents a distraction from the selling message, the promotion may actually aid in acceptance of a brand for which consumers have negative attitudes, or inhibit increased acceptance of an already favorable brand. Information overload The difficulty in processing the incoming information from a message has been thought to influence the capacity of the receiver to immediately engage in cognitive response, thereby reducing the amount of counterarguing (Wright, 1973). Respondent processing ability could, in turn, depend on the message quality, the message layout or presentation approach, or the mode of presentation. 89 Wright (1973) experimented with message modality which directly relates to information load and found counterar- gument to be the dominant mediating response among those who received an audio version of the message versus those who received a print version of the same message. Although distraction was not tested against this intermedia compari- son, the implications are that in a situation which does not allow leisurely processing of data, persons rely on arguments against the message rather than supportive feelings. In the case where a coupon or price promotion within an advertisement adds to the information load and deters processing, it is a good assumption that counterarguments would dominate and the outcome would be less favorable brand attitudes. Furthermore, sales promotion would receive negative ad evaluations as compared to advertising. Summary: Putting it all in Perspective Several perspectives have been presented which offer potential insight into the effects of sales promotion on brand image. These theories, however, are not applicable to the same stages of the consumer choice process involved in a purchase event. Cognitive dissonance, attribution theory, and self perception theory all are predicated on internal processes, or external interpretations of internal processes, which take place pips; the purchase of a product due to the conditions of sale. These influence repeat purchase of the same brand. 90 Post-purchase dissonance and attribution processes are not directly observable and are often unconscious to the individual. Therefore, it is very difficult to isolate, measure and study these internal processes in a controlled way. They must be surmised. Moreover, since both theories predict the same outcome in a purchase situation, distinguishing as to whether dissonance or attribution is operating is practically impossible. The few researchers who have looked at one or the other of these phenomena usually conclude by saying that the results could also be explained in terms of the other theory. In addition, dissonance is generated due to the purchase task itself and the competitive environment; the type and extent of attribution depends on what other external/internal facilitating factors are present at the time of purchase which might discount the purchase motivation. For reasons of external validity, then, a real purchase environment involving longitudinal purchasing patterns is necessary to study these post-purchase theories. Price perception theories work to explain both purchase and repeat purchase behavior, specifically at the point-of- purchase when price becomes a key consideration. Adaptation is a time-based variable and the establishment of acceptable or standard price points or ranges require a longitudinal design in a realistic competitive environment with controlled price conditions and discounting procedures. The decision not to study these point-of—purchase and post-purchase theories was twofold. First, the difficulty 91 and expense in attaining an acceptable level of internal control over measurement within an externally valid field setting is almost prohibitive. Second, the sales orientation does not allow careful investigation of the communication effects of advertising variables, like the inclusion of a coupon, on attitudes before purchase. The learning and information processing theories are much more applicable to a non-purchase situation. These theories involve the learning and processing of information prior to purchase. The intermediate effects influenced at this stage are important since exposure to an advertising message constitutes an important learning experience which influences the brand attitudes influencing behavior. These criterion variables are not subconscious; therefore, their magnitude can be measured directly by self report. The degree of learning, affect and purchase motivation, as well as any subvocal cognitive responses, can be measured in a laboratory more precisely and at less cost without the complication of a purchase situation. At this exploratory stage in sales promotion response research, this precision is important. In the lab, it is possible to integrate and manipulate non-purchase variables such as advertising and repetition, and hold other marketing variables constant, such as competitive clutter. Since purchase is based not only on attitudinal predisposition but also on other uncontrollable variables such as product availability, in-store display and shelf facing, disposable income, it is more sensible to study 92 TABLE 2.3 THEORETICAL PERSPECTIVES FOR SALES PROMOTION EFFECTS Theory Role of a coupon Effects POST PURCHASE SITUATION Dissonance Theory Helps justify purchase, Weakened brand attitudes reducing post-purchase less probability of future dissonance purchase Attribution Theory Reduces price relative Perceptions of inferior to non-discounted quality competitors Reduces price as a Parity perceptions and retaliatory measure less image differentiation Self Perception Acts as extrinsic Weakened brand.image and Theory attribute and discounts less probability of future intrinsic attributes purchase POINTHOF-PURCEASE SITUATION Price/Quality Increases salience of Weakened brand image extrinsic price cue Adaptation Theory Adjusts price point Less probability of future downward purchase at regular price Assimilation- Narrows acceptable Less probability of future Contrast Theory price range and adjusts purchase at regular price range downward 93 TABLE 2.3 (Continued) Tummy Rohscfl’acxwmon Effiufis Behawuuadjhmundng ‘flumry Cognitive Learning ‘flmnry Infinmmthm: Processing W Rmuudslmuchmue or trial Attracts attention to adwntimmmmt Establishes hard sell creative orientation Initiates shallow processing Owniommlsemuuy input Distracts from brand and inhibits counter- mnndng Establishes habitual purchase without attitude ckwehmmmmt Immmmmmilemnflngtmmlmmne favorable attitudes Increased initial learning, decreasing over repetition Less recall, weakened brand :hmge Less recall, weakens brand hump Increased yielding to unfavorable brand or ad; Decreased yielding to favorable brand or ad the precursors to behavior under controllable conditions so that an initial understanding of the cause-and-effect relationship can be developed. Table 2.3 categorizes the various theories according to their appropriateness at the different stages of a brand choice situation. basic role of a coupon is thought to be, This table states for each theory what the along with the hypothesized effect on the relevant criterion variables. 94 Exposure Freguency In the previous discussion of learning, information processing and cognitive response, repetition of exposure to a message or an advertisement seemed to have an important influence on the outcome measures. This shouldn’t be surprising. Repetition is a basic tenet of learning, and wearout of affect has long been a concern of advertisers. Exactly how and why repeated exposure to the same message mediates attitudinal effects is subject to many hypotheses, some of which have been presented. What is significant, is that over repeated exposures, effectiveness measures at all levels often show reversals in direction of impact, that the degree and direction of impact varies according to criterion measure, and that diminishing returns and wearout have been shown both in the laboratory and in the field. For all these reasons, repetition becomes an important variable in any study of advertising or communication effects. There is yet another reason to include repetition in a study of communication effects. Communication research has shown that initial message source effects decay over time either increasing the effectiveness of a message with negative source effects, or reducing the effectiveness of a credible message. This has been dubbed the "sleeper effect." The sleeper effect would imply that any negative effects of a deal on a consumer might become disassociated with the brand over time and, therefore, not damage brand image and loyalty (Aaker, 1972). This proposition, however, may not be appropriate in repetitive stimulus conditions. Weinberger 95 (1961) has shown that repeated exposure to the negative or positive source reinforces the original attitude toward the message. If an advertisement emphasizing a promotional deal creates negative attitudes initially and is repeated often, the sleeper phenomenon would probably pp; dilute the negative source effect as Aaker hypothesizes. Brezen (1985) and Sawyer (1974) have reviewed the repetition research in detail and the most relevant of these studies have already been presented; therefore, this literature will not be reviewed here further. However, some general conclusions about repetition effects and the issues surrounding exposure control deserve elaboration. General Repetition Effects Recognition has usually been found to be greater than recall, yet the most memorable ads are not necessarily the one which elicit higher level attitudinal effects. Recall appears to be more volatile than attitudes and the higher level motivational measures like purchase intent, with both steeper generation and decay. Grass and Wallace (1972) found attitudes to peak at a higher exposure level than recall and to decay at a slower rate. This may suggest that different frequency levels are required to effectively enhance and sustain different levels of response. Most repetition research showed effect peaking at two to four exposures and all the studies showed diminishing returns. However, it is difficult to generalize too much 96 because each study was executed under varying conditions, using varying effectiveness measures. Repetition effects interacted with creative strategies, media scheduling strategies, media mix, and in some cases, product characteristics. Exposure Control Exposure has been a badly mistreated concept. In evaluating the "exposure" in the repetition literature, a variety of exposure definitions have been used and confused, including vehicle exposure, ad exposure and advertising communication. The "opportunity to be exposed" (OTS) is common way to define exposure in field studies where it is difficult to know exactly who is viewing or reading. Since the exact exposure definition implemented will affect the results of a study, this is a serious semantic error which makes it difficult to unify the literature. The decision of which exposure treatment to use is an issue over the trade— offs associated with the internal control and the external validity of the specific Operationalization. 1 Laboratory experiments assure tight control and it is in this instance that actual exposure can be studied. However, the demand characteristic of forced exposure is not reflective of the way advertising is actually distributed and viewed and therefore, impairs external validity to some degree. In field experiments, the best exposure measure possible is OTS-—the probability of exposure. OTS does not indicate actual exposure, since an individual may not have been in the 97 room during the television commercial breaks, or may have skipped through a magazine without seeing an advertisement. It is much harder to ascertain which individuals have seen the ad exactly once, twice, etc. Field experimentation does rate much higher on replicating the real environment, however. Measurement over a longer time period and over a greater number of exposures is possible than when a subject is confined in a laboratory setting. Simulation of the media is not necessary as it is in a lab and the outside environment is allowed to vie for consumer attention and mediate the advertising impact realistically. With this, of course, control is diminished and measurement is much less precise, hence it is difficult to gain an understanding or prediction of individual advertising response. Furthermore, the expense of a large field study has in the past been a major deterrent to this type of research. Ray and Sawyer (1971, p. 21) have evaluated the various approaches used in repetitive advertising research and this is shown in Table 2.4. 98 TABLE 2.4 APPROACHES IN REPETITION RESEARCH Exposure Typical Relative Type of Study Control Measure Repetition Cost Econometric or cam- Number of paign insertion Very weak Sales insertions or Moderate ¢mqmmdrun1s Advertising survey Claimed Advertising or panel purchase Weak Purchases campaign High Advertising field Single or Very experiment Moderate Sales multiple ads High Advertising lab experiment Absolute Attitudes Single ads Low Research Questions and Hypotheses The major research questions are: (1) (2) (3) (4) Does exposure to print advertising with coupons affect brand attitudes differently than exposure to the same advertising without coupons? Does repeated exposure to advertising with coupons as opposed to those without, influence the outcome? Does deal proneness mediate attention to coupons? How do attitudes differ for a promotion- advertising mix as compared to promotion (coupons) or advertising alone? Each of these will be addressed in detail and the hypotheses for each outlined. The first research question deals with the differences between advertising and sales promotion (advertising with a 99 coupon) on the cognitive, affective, and conative dimensions of attitude development, or brand and advertising awareness, brand evaluation, and purchase intention, respectively. The effects on each criterion variable are discussed. Coupons have been shown to increase attention levels to advertising (Bowman, 1985) and if a coupon represents an attention—grabbing element, or a hard sell approach, these approaches have been found to have higher recall than non— grabber ads or soft sell ads, respectively (Ray and Sawyer, 1971; Silk and Vavra, 1974). Yet, coupons can be hypothesized to distract from the main message components, to overload sensory input reducing the amount of brand information that can be processed, or to act as "shallow" processing cues which would reduce memory transfer. These latter information processing theories strongly support the notion that sales promotion would elicit less recall than advertising. However, a two-tailed hypothesis is stated at this exploratory stage of study. Hypothesis 1: Brand and advertisingrelated recall (learning) will be significantly different (lower) for ads with coupons than for ads without coupons. According to cognitive learning theory, increases in learning are accompanied by increases in affect; therefore attitudinal differences between sales promotion and advertising would parallel those differences found in recall levels. 100 Hypothesis 2: Brand evaluation will be significantly different for ads with coupons than for ads without coupons. A coupon could serve to cheapen a brand’s image or increase perceptions of its value. Purchase intent would be strengthened if a brand was perceived as an excellent value for the money and weakened if the savings of a coupon reflected negatively on value. Learning theory, however, would state that purchase motivation is dependent on the magnitude of favorable attitudes. Hypothesis 3: Brand value and purchase intention will be different for those subjects who were exposed to an ad with a coupon versus those who were exposed to an ad without the coupon. Studies have shown overload to increase the negative response to the message which is overloaded with information. Wright (1973) showed that people tend to rely on counterarguments more than support arguments in such a situation. If a coupon is hypothesized to add to the clutter of an ad and increase information overload, the overall valence index for all thoughts, as well as the ad evaluation index and product-related counterargumentation should be more negative for the sales promotion condition. If a coupon, on the other hand, does not represent clutter, but instead distracts from the other message components, normal negative coEnitive response will be decreased. Overload and 101 distraction concepts, then predict differing results. Again, two-tailed hypotheses are most appropriate. These are stated for all three cognitive response indicators. Hypothggis 4: Overall cognitive response to an advertisement with a coupon will be different in valence from an advertisement without a coupon. Hypothesis 5: Ad—related cognitive response to an advertisement with a coupon will be different in valence from an advertisement without a coupon. Hypothesis 6: Product-related cognitive response to an advertisement with a coupon will be different in valence from an advertisement without a coupon. Information processing indicates that selective attention will affect what information is transferred into LTS. One factor especially relevant to the study of sales promotion, deal proneness, might influence what is attended to within an ad. On the assumption that deal prone persons, or those who use coupons, will be more interested in a coupon than nondeal prone persons, it is hypothesized that: Hypothesis 7: Those who use coupons will recall the presence of a coupon in the advertising at a different rate (greater) than those who do not use coupons. The interaction of advertising and sales promotion in a campaign is of interest to advertisers because rarely is sales promotion used without advertising and vice versa. 102 Strang (1975) has reported that a number of companies found advertising and promotion "can interact synergistically to produce a higher level of sales than comparable investments in either advertising or promotion alone" (p. 14). This synergy has not been tested at an individual level in a controlled exposure situation. However, it may stand to reason that exposure to a mix of regular advertisements and similar advertisements with coupons, would sustain attention longer due to the execution variations, increase learning as compared to exposure to the promotion condition alone (less distraction over all exposures), and increase value as compared to advertising alone. Therefore, it is hypothesized that: Hypothesis 8: Exposure to a mixed condition of both advertisements with and without coupons will affect all criterion variables differently than exposure to either alone at equal exposure levels. Repeated exposure to a message increases the opportunity to attend to, comprehend, and learn the content of that message. According to the learning hierarchy of effects paradigm, increased learning increases the probability of favorable attitudes and purchase intentions. However, the twofactor theorem also suggests that at some point over higher levels of repetition, tedium or inattention will cause wearout or negative impact on learning which in turn will affect higher levels of attitude development. This would also be reflected in overall cognitive response. The 103 hypotheses regarding repetition, then, are stated as two- tailed at this exploratory stage and are formally presented below. Hypothesis 9: Brand and advertisingrelated recall will differ over repetition levels. Hypothesis 10: Brand evaluation will differ over repetition levels. Hypothesis 11: Brand value and purchase intentions will vary at different exposure levels. Hypothesis 12: Cognitive response will vary at different exposure levels. CHAPTER III METHODOLOGY Research Design A laboratory experiment was designed for the task of testing different advertising—promotion conditions over different repetition levels as outlined in the hypotheses in Chapter II. Where repetition is involved, the ”opportunity to be exposed" a given number of times to an advertisement is not sufficient to speculate with any precision the response function shape. Since differential slopes or attrition of brand attitudes between advertising and sales promotion is a key concern here, precision is important. This eliminates field experimentation as a viable design alternative. Even a direct targetable medium such as direct mail, used by Strong 1972), has less-than-the—desired control given the clutter of direct mail solicitation and the increasing likelihood that unsolicited mail will not get opened. The self-selection of individuals in exposing themselves to such promotional activity in-home is a problem with direct mail field experi— mentation since deal proneness is also a variable under study in this research. Lab exposure also allows for constant exposure across all promotional pieces and control over the amount of competitive advertising clutter. 104 105 Three repetition levels (1, 2, 3) were crossed with three promotional conditions (advertising alone, promotion alone, or a mix of both) such that nine experimental conditions and one control condition were possible for a brand, as shown in Table 3.1. TABLE 3.1 EXPERIMENTAL CONDITIONS Advertisement Advertisement With Coupon Withopt Coupon 1 0 2 0 3 0 0 l 0 2 0 3 l l l 2 2 l 0 (Control) 0 Although no explicit hypotheses regarding product category or brand position were suggested in Chapter II, differences between categories most certainly exist. There- fore, to avoid the bias of any one product category, eight brands from eight non-competitive categories were selected for study. Each of ten experimental groups served as a control for one brand and was exposed to a different promo— tional condition for each of the remaining seven brands. Because a complete factorial design would require ten cells for each of eight product category advertisements or an unwieldy 80 experimental cells, an incomplete or fractional factorial design with repeated measures across product Groups A P 106 categories was used. The design falls short of a true latin square because of some confounding due to the rectangular shape resulting from only eight test product categories and ten promotion—repetition levels. Although this limits the statistical power of the design in reducing within group error, individual product category analyses at this explora- tory stage were thought to be more appropriate given the obvious creative execution and category differences. The design is diagrammed in Figure 3.1. Brands A B C D E F - E E Total Test Ads 9 m > m > w > w > m m w > 'o (l) (2) (3) (4) (5) (6) (7) (8) (9) (10) l6 16 16 16 16 16 16 16 16 16 2 O 3 0 1 O 1 1 O 2 HwHOv—‘NNOOO OOr—Hv—OLoONN wHOr—‘NNOOHO owwwowONNo HOv—‘NNOOHOCJ HwI-‘Or-‘ONNOO OONNr—Ov—‘Omr— OHOU’ONNOr—‘H v—NNOOOD—‘wt—O HOwONNOOv-‘H NNOOHowv—OH OwONNOOr-‘HH NOOHOwr—‘Ov-‘N HONNOOer-‘O v-‘Ov-‘OUJHOONN Fig. 3.1 Repeated Measures Design Across Eight Brands Procedure A total of 16 test advertisements across the eight product categories, randomly intermixed with 34 other non— competitive ads, were shown to each group of subjects. To allow for equal exposure time to all ads by all subjects, the 107 print advertisements were transformed into color slides and randomly sequenced into a slide presentation. To avoid any primacy or recency effects, no test ad appeared within the first or last three frames of the presentation, and to reduce the possibility of any excess synergism due to repetition, no two test ads for one brand appeared closer than six frames from one another. Furthermore, no ad repeated itself a second time until at least halfway through the 50-frame sequence to prevent subjects from selectively exposing them- selves to repeated ads. To minimize order effects, 16 test ad "slots" were randomly selected initially among the 50 frames within the stated guidelines. For each experimental group, the 16 test advertisements varied according to the promotion-repetition conditions for the brands in the particular group and were randomly assigned one of the 16 designated test ad positions. The remaining 34 ads were randomly assigned positions and both the ads and their positions remained constant across all groups. In this way, ten different slide trays were assembled for the ten experimental groups which could be repeatedly administered to attain enough subjects per cell. The experimental sessions took place over a two month period from early August through early October. The sessions were conducted personally in the same classroom to avoid any environmental or procedural variation. The selected classroom was relatively small so as to limit seating capacity to approximately fifteen people and to assure sufficiently clear viewing range of the screen from every 108 seat in the room. Large individual tables allowed ample space and comfort for unhampered and relaxed writing. Each experimental session lasted approximately one hour beginning with a brief disguised introduction to the study purpose and procedure with time to clarify any questions. All subjects completed consent forms confirming the voluntary nature of their participation and then watched the slide presentation. The projected image of the ads was kept constant in size from session to session and was sufficiently large so that much of the body copy, in addition to the headlines and visuals of the print advertisements was very legible. Allowing for ten-second exposure to each of the 50 print slides, the whole viewing process took only about eight minutes, after which self-administered questionnaires were completed by the subjects. Participants were told not to reveal the content of any of the questions or slides to others who might potentially participate. However, because of the disguise, any talk most likely would have been harmless. In feet, through later contact with many of the participants, none guessed or indicated any knowledge of the true nature of the experiment. Disguise of the Experiment Experimental subjects were told that they were partici- pating in a study about advertising campaigns and their effectiveness. Specifically, they were told to watch all the advertisements and pick out any campaigns among them. A campaign was defined as a series of different ads for a 109 particular brand or branded line of products where, although the ads are similar, each one is different from the other in some aspect. In other words, several repetitions of the same ad did not constitute a campaign. This disguise was used for several reasons. First, the campaign concept inherently involves repetition; hence, repetition of a test advertisement would not arouse suspi- cions of the true nature of the experimental manipulations, therefore eliminating any bias that might result. Second, since the slides were randomly ordered so that none of the campaigns or test ads repeated too early in the sequence, attentiveness to all the advertisements throughout was main- tained. Third, this disguise did not focus attention to specific ad components or characterizations and away from the variations in layout and promotion as another disguise might. For instance, using the disguise of evaluating women’s role portrayals in advertising might result in subjects focusing analysis on character situations and completely overlooking any creative or message characteristics. On the other hand, the campaign introduction did not mention any specific varia— tions to look for among advertisements in a campaign which might draw attention directly $9 the differences in cell conditions. Three campaigns in all were included among the ads shown to subjects. Each campaign consisted of four individual advertisements which could be immediately recognized as belonging together in a campaign, although they were obviously different. The first campaign, for Napier jewelry, 110 showed different women wearing different costume jewelry pieces with similar headlines in a consistent creative layout. The second campaign was for Spanish olives and showed olives being served on various food dishes in common color schemes and uniform headline themes. The third campaign sported different Nieman Marcus stores in different cities with different designer names but in a creative style that was unique. None of the campaigns were in product categories competitive with any of the test ad categories. The latter two campaigns were relatively obscure in the sense that they had not appeared in the print media frequently. This made the task of identifying the campaigns more challenging and increased attentiveness to that task. The 16 campaign ads are shown in Appendix I. Immediately following the slide presentation but preceding any questions about the specific test ads under scrutiny, questions concerning the campaigns were adminis— tered. These asked subjects to describe each ad within up to three campaigns that they could remember. This questioning not only supported the disguise, but also allowed enough time to pass between advertising exposure and post—testing so that recency effects of the test advertisements would be erased. Since many of the test advertisements were repeated two to three times within the presentation, asking questions about these ads along with questions about the campaign sequences, was not as suspicious as it might otherwise have been. Furthermore, questions regarding purchase intent and brand evaluation were appropriate for a study measuring lll campaign "effectiveness." Throughout the question sequences on the different criterion measures, one of the campaign brands, Spanish olives, was always represented, again to carry through the disguise. Sample Women are major purchasers of health and beauty aid products as well as package goods items and are more likely to use coupon promotions than men (Nielsen, 1985). Subsequently, women are a primary target for advertising and promotions by advertisers. Therefore, a convenience sample, consisting of 393 females, was recruited on a voluntary basis from campus organizations and classes at Michigan State University. This sample, of course, represents the aggrega— tion of subjects from the ten experimental conditions, each of which contained approximately 35 subjects. The exact cell counts are shown in Table 3.2. TABLE 3.2 EXPERIMENTAL CELL SIZES Experimental Group Number Group 1 39 Group 2 41 Group 3 33 Group 4 40 Group 5 33 Group 6 43 Group 7 39 Group 8 45 Group 9 40 Group 10 39 112 The target age range for the sample subjects was 18 to 34 years, the primary audience demographics of the selected test product categories and the print ad vehicles from which the test advertisements were taken. Any subjects indicating that they did not purchase any of their own personal and food items, or who did used less than four of the eight selected test product categories, were excluded from the sample. Selection and Preparation of Stimuli Mpterigl Using fictitious brands or developing dummy advertising for existing brands in order to control for prior exposure to the test advertising was considered. However, due to time and cost constraints, these alternatives were discarded in lieu of using existing advertising for already existing brands. All test advertisements were one-page ads currently running in and taken from women’s fashion magazines such as lggpg, Madegoiselle, glpmour, Self, Cosmopolitan, and Bazaar. Since several manufacturers have advertised both with and without promotion by simply adding a coupon to the original print advertisement, there was no need to design fictitious coupons or fictitious ads for this experiment. The advertisements were matched for the promotional manipulations so that the promotional advertisement-~the advertisement with the coupon-—was virtually identical to that without the coupon-~the non—promotional test advertisement. Only for one test brand, Crystal Light, was it difficult to find an exact replica of the two versions. To 113 remedy this, a coupon from one ad was simply superimposed over a similar ad without the promotion in such a way that an identical version was created. However, even in this situation, the coupon was an existing promotion and the newly created ad was very similar to those in the media already. Because all coupon redemption values were part of regularly scheduled promotional programs by advertisers, the values all met industry standards. Several criteria were used in selecting the test ads as well as the filler ads for the experiment. Of course, availability of advertising both with and without promotion for the brands and the appropriateness of the category for the sample were foremost on the list of criteria. All categories or brands used as either test or control ads are purchased more frequently by the 18 to 34 year old age group than all women according to industry sources. Categories were limited to those with short purchase cycles and frequent usage which would be more appropriate for short-term coupon incentives. These were primarily health and beauty aids and package goods products. In fact, a Nielsen study (1985) indicates that use of coupon for toiletry items has risen from 73x in 1975 to 88X in 1984, while 98% of shoppers use coupons for food items. Categories were also those which were appropriate for the late summer or early fall seasons. All advertisements were chosen with their seasonal flexibility in mind. Given the data collection period, ads were selected which could be held over into mid—October if necessary. Bathing suits and suntan products, blue jeans, instance, 114 perfume and nail polish. as well as winter ski fashions, for were substituted with transitional products such as Because past research has shown that variations occur according to the competitive position of a brand in the marketplace, some effort was made to include brands which were dominant in their category (40% market share or more) along with brands which were competitive (holding 20% market share or less). No ads for competitive brands within a product category were included in the presentation in order to eliminate competition effects. The categories and brands selected are shown in Table 3.3 along with their individual ad characteristics. The actual reproduced advertisements for both experimental conditions are shown in Appendix II. TEST BRANDS AND TABLE 3.3 ADVERTISEMENT CHARACTERISTICS Product Category 1 . Mascara DON OS’I-b 7. 8. . Hair Remover Nail Polish Remover Face & Body Soap Powdered Drink Mix Demkuant Headache Remedy Sanitary Napkins m Maybelline Dial-a—Lash Neet Cutex Tone Crystal Light Mennon’s Lady Speed Stick Advil Nawireedmn MiFonmn 4-color 4-color Black & 4-color 4~color 4-color 4-color 4-color Whfim Cmumn Face Value $.50 .30 .15 .20 .30 .35 .50 .50 115 Operationalizatiop of the Experimental Variables The development of the independent variable manipula- tions and the operationalizations of the dependent criterion measures, as well as for other measured variables such as product usage, brand loyalty, and deal proneness which might be used for comparison purposes, are now explained. The exact question wording and answer categories can be found in the study questionnaire in Appendix III. Independent variables The independent variables under study--repetition and promotion mix-~and their operationalziation in this design have already been discussed. The development of these opera- tionalizations, however, deserves elaboration. Repetition In past repetition research, the tested exposure levels have varied considerably from study to study. Table 3.4 gives the number of repetitions tested for several studies. As clearly evident from this brief review, up to 24 controlled exposures in a laboratory for a brand have been tested. Several studies only looked for and found differences between one and two exposures. In studies where researchers have tested some combination of up to four, five or six, they have often found measured response to peak at two to three exposures or at least plateau after this point. This study focused on one, two and three exposures for two reasons. First, past studies have shown this number 116 TABLE 3.4 NUMBER OF EXPOSURES TESTED IN PAST FREQUENCY RESEARCH Article Frequency Levelg Weinberger (1961) Silk and Vavra (1974) Saegert and Young (1982) Singh and Rothschild (1983) Grass (1968 ) Belch and Belch (1984) Belch (1982) Stephens and Warren (1984) Ray and Sawyer (1971) Sawyer (1973) LoScuito (1967—68) Cacciopo and Petty (1980) Calder and Sternthal (1980) 18 Craig, Sternthal and Leavitt (1976) 7, 12, 21 6, 12, 18, 24 NNNNN 9b NEON“) 0300000050 000000 9P0 Ol (”51' OIOI 05050) HHI—‘t—‘HHH .. vs U . . O U U . n—u—u—u-I . O . .S" sufficient to capture initial fluctuations in response generation, diminished returns and satiation. Second, aggregated over eight test brands, more than three exposures per brand would add to the complexity of the design and increase subject reactance to the experiment. Promotion manipulation Up to three repetitions of advertising alone and promo— tion alone allowed direct comparison of these two conditions. The inclusion of several advertising and promotion mix conditions stems from econometric research which has shown that a mix of advertising and promotion is better than either used alone. However, the best ratio of one to the other is tenuous. Hence, an equal split was tested (one exposure of each) against an advertising heavy and a promotion heavy mix (a ratio of two advertisements without coupons or two ads 117 with coupons versus only one of the opposite condition, respectively). Dependent Variables A variety of criterion measures were used in this study, although the primary thrust of the research involves effects on brand image. There are several reasons for this. First if image is equated to attitudes, the traditional attitude definition has three components: cognitive, affective and conative. To measure the three levels of attitude develop- ment, then, requires measurement of more than just the affective aspect. Second, Strang and Gardner’s portfolio study (1983), similar to this study in many respects, including its intent to look at brand image, investigated brand evaluation via several bipolar adjective rating scales and found nothing. This leads one to ask if perhaps, in a short term laboratory experiment, it is just as important, if not more important to concentrate on differences in response to promotion versus advertising at lower levels of response such as recall. Third, in repetition research it is not uncommon to find that response varies greatly according to the criterion variables measured. For instance, Sawyer (1973) found that the most memorable ads were not the ones which most effectively increased purchase intent. Ray and Sawyer (1971) found response to different criterion measures often crossed back and forth over the span of even six exposures. They also found interactions to occur between other predictor 118 variables such as advertising format and criterion measures with increased frequency. Silk and Vavra (1974) had similar results. Furthermore, to test the merit of the theoretical propo- sitions discussed in Chapter II, it is possible cognitive response and recall levels might offer insights beyond what a direct magnitude scale of evaluation or purchase intent might reveal. For these reasons, the primary criterion measures measured were (1) unaided brand recall, (2) brand aided advertising related recall, (3) cognitive response, (4) brand evaluation, and (5) purchase intent. Each measure, its specific Operationalization and coding scheme are now described. Brand Recall A very simple category-only-cued brand recall question was the first measure to follow the distraction questions regarding the campaigns. The question asked the subject to write down any brands in each of the product categories listed for which she could remember having seen advertising in the previous presentation. Three spaces were provided for each product category cue, although only one brand for each was shown. Subjects were instructed to leave the space blank if they could not remember any ad for a category. Each category then was coded as no recall, correct brand recall, or incorrect brand recall. If a subject mentioned more than one brand, only the first brand mentioned was 119 coded. A description rather than a brand name, such as, "the one with the dots," was interpreted as no recall. For mascara, both Maybelline and Dial-a-Lash were accepted and for deodorant, both Mennon and Lady Speed Stick were accepted. Misspellings were ignored and coded as correct unless the spelling was too mangled for clear interpretation. Since Advil and Faberge are relatively new products, recall of these unfamiliar brand names was often hard to decipher. Advertising relgted recall 52g coupon gention Subjects were given a brand name cue for all eight test brands whether shown in that session or not and asked to record in detail as many different things they could remember about what the ad showed or said. Following this, for each brand the questionnaire prompted the women to list any thoughts they could remember having had while viewing the advertisement, or their response to what the advertisement showed or said. These open-ended answers were coded into one of four categories for ad recall. If there was no response at all or if the description was completely incorrect, this was coded as no recall. Depending on the thoroughness and accuracy of a written response, ad recall was coded into one of the three remaining recall levels. The lowest recall level represented answers which were so vague or general in nature that they could be interpreted as describing other ads in addition to the test ad, i.e., ”it showed a model with a picture of the product." The second level of recall represented related or 120 qualified recall, meaning the subject had mentioned some specific detail that indicated she had certainly seen the test advertisement under consideration. For instance, "there were two girls back-to—back" or "it was black and white and showed a graph." The fully qualified recall category required a subject not only to describe the ad to some degree, but also to correctly relate the main copy point or slogan. Typical responses for two product categories coded at each recall level are shown in Table 3.5. TABLE 3.5 AIDED RECALL CODING EXAMPLES Cutex No Recall: Don’t remember; A large nail with red polish Vague Recall: A picture of the product Qualified Recall: A large bottle in black and white with a graph Fully Qualified: A bottle and graph showing research which proves Cutex makes your nails stronger. Tone Soap No Recall: A women running through a field Vague Recall: Showed a model using soap Qualified Recall: A woman half in and half out of the shower Fully Qualified: A woman in the shower using Tone with the words Bye Bye Dry 121 Because every test group saw a different condition, only some of which included couponed ads, mention of a coupon correctly or not was not considered when evaluating recall. However, since the inclusion of a coupon was hypothesized to be a distraction from recall of other advertising components, whether or not a coupon was specifically remembered from the advertisements by a subject was noted and coded separately. All coding of recall and coupon mentions was conducted without any knowledge of the actual cell condition for a particular respondents in order to eliminate any subjectivity or unconscious bias in the coding process. By recoding approximately one-third of the questionnaires in a random manner among the ten cell conditions, coding consistency and reliability was thoroughly checked. Cognitive response To prepare the cognitive response data, ten categories were defined to code the response to the ads. These ten categories are listed in Table 3.6. Studies in the past have varied in the measurement of the various thoughts collected from this type of question. Very broad general categorization schemes, such as the favorable and unfavorable thoughts used by Calder and Sternthal (1980) and Cacioppo and Petty (1978), have been used. However, these thoughts were specifically directed toward the message and/or product either by direc— tions or design. Wright (1973) labelled these thoughts as counterarguments and support arguments, and added source derogations. Belch (1980) compiled a coding system consis— 122 ting of 15 different cognitive response categories which expanded on these broader schemes and included advertising— specific variables. TABLE 3.6 COGNITIVE RESPONSE CATEGORIES l. Counterarguments 2 Support arguments 3 Disaffirmations 4 Affirmations 5. Negative Ad Evaluations 6. Positive Ad Evaluations 7 Irrelevant thoughts 8 Neutral thoughts 9 Repetition-oriented thoughts 0 1 . Non-product user The ten coding categories used in this study are derived from Belch’s coding categorization. Essentially they consist of message/product related thoughts, advertising execution related comments, and irrelevant thoughts. Message or product related thoughts were divided into two sets. Support arguments and counterarguments were those statements about the product or message content which expressed belief or disbelief with the message, or favorable or unfavorable thoughts related to the use of the product or its attributes. Affirmation and disaffirmation of the product, on the other hand, included simple statements of like or dislike for the product or brand not accompanied by any explanation. Comments directed toward the execution of the adverti— sement rather than the advertised product, were coded as positive or negative, and included statements about the choice of the model, the layout and clutter of the piece, or 123 the advertising approach used, i.e. comparison advertising. Although Belch also had categories for source bolstering or source derogation which covered any comment regarding the advertiser and the advertiser approach, these were combined in the study with those thoughts regarding the execution. Completely irrelevant or completely neutral thoughts to either the message or the advertising were coded separately. Examples are "the ad reminded me I have aerobics today" and "it just showed the product being used," respectively. Two last categories were added to cover instances where the subject (1) stated that she did not use the product, and (2) specifically commented on the repetition in the design. Table 3.7 shows examples of these cognitive response variables for two product categories. The precise opera— tional definitions of these ten cognitive response codes, adapted from Belch’s list, are found in Appendix IV. Wright wrote about several problems associated with the monitoring of cognitive response. One concern over measuring the thought process was the extent to which the thoughts recorded by the individual are thoughts which actually 124 TABLE 3.7 COGNITIVE RESPONSE CODING EXAMPLES Cutex Counterargument: I don’t believe any remover can actually strengthen your nails. Support argument: My nails always split and maybe Cutex could help. Disaffirmation: I don’t like Cutex. Affirmation: I use Cutex and like it. Irrelevant: I wonder where my nail polish remover is? Negative Ad This ad is boring in black and white and looks too Evaluation: scientific with the graph. Positive Ad I like the fact that they showed Evaluation: a graph proving their point. Repetition: This ad became boring after the third time. Neutral: Nothing special. Didn’t attract my attention. Don’t purchase: I don’t use nail polish so I don’t pay attention to these ads. Ma elline Dial—a—Lash Counterargument: I think this is a gimmick. If you want thicker lashes, just apply several coats; I’ve used this and it doesn’t work like they say it does—~it clumps. Support argument: It’s a neat idea because you can use the same product for any mood or occasion; Sure sounds like a versa- tile product. Disaffirmation: I would never buy this product. Affirmation: Good brand. Negative Ad Too cluttered; This is a stupid way to advertise this Evaluation: product--you can’t see their eyes. Positive Ad I like the way they showed the two different women Evaluation: with active lifestyles; Pretty models with pretty smiles. Irrelevant: Reminds me of my sister; I hate hats! Repetition: I’m sick of seeing this ad. Neutral: Just another mascara ad... 125 occurred during exposure. Of course, having subjects actually record thoughts at the time of exposure is one option, although it is impractical for many experiments such as this one. Recording thoughts after the fact has been found to generate slightly more counterarguments as compared to during—exposure thoughts (Roberts and Maccoby, 1973). This same concern applies to situations where the researcher unintentionally promotes the cognitive response measurement or one type of cognitive response over another. Subjects aware of the researcher’s interest in the cognitive thoughts might artificially inflate their general level of response activity. If directed to elaborate on specific cognitive response types, such as counterarguments, subjects could address these particular thoughts at the expense of other response sets. To reduce these effects in this study, no mention was made prior to advertising exposure of the interest in cogni— tive response and the question which addressed this particu- lar variable did not specify any one type of response over any other. Other studies have imposed strict time limits on recording response to each message. Because the posttest questions were self-administered, this was not possible. Furthermore, since everyone does not write at the same speed, the idea of introducing an arbitrary time limit on answering these questions was discarded as it might hamper some people while encouraging others to fill unused time by generating fictitious comments. 126 Brand evaluation Two measures of brand evaluation were used. In lieu of asking an entire series of appropriate adjective rating scales for each test brand which could be summated for an aggregate measure of evaluation, a single five-point compara- tive rating scale was implemented. This scale directed the subject to place each brand within an ordinal hierarchy ranging from "One of the Worst" to "The best in its category" and is similar to the brand evaluation measure used by Strong in his repetition research (1972). In addition, a ten-point rating scale of value for the money, ranging from "Extremely Poor Value" to "Extremely Good Value" was included since value was considered to be one adjective scale which might be directly affected by the couponing conditions. These two scales can be found in Appendix III. Purchase intent Purchase intent was measured via another ten-point scale asking each subject to indicate how certain she was to buy the brand on her next purchase of the corresponding test product category. Other Experimental Variables Product usage Because experience with a category or relevance of a category is apt to influence prior familiarity with brand names and attributes and, therefore, the selective exposure to advertising in the category, an estimate of product use was included in the questionnaire. First, subjects were 127 asked to indicate how frequently they purchased the product and second, they were asked if they had purchased any of the product in the last six months. Any subject who did not use or purchase at least four of the eight test categories was excluded from the study. gppnd loyglty Brand loyalty is an important covariate in this study of promotion since prior research has shown brand switching or low brand loyalty to correlate with susceptibility to promo- tional dealing, as discussed previously in Chapter II. Table 2.2 has already referenced the means by which this concept has been measured. As reviewed in Chapter II, four basic operationalizations have been used: (1) a nominal categorization scheme based upon the pattern or sequence of brand choices over time, (2) a ratio-level degree of loyalty representing the proportion of total purchases devoted to the single most purchased brand in the time period, (3) a ratio- level count equaling the total number of brands purchased over the period, and (4) a dichotomous loyal/not loyal classification based on stability over time as determined by comparing the "favorite" brand of two different time periods. All of these measures were developed for diary panel data or experimental data in situations where subjects actually "purchased" or selected sequences of a brand in the real world or a laboratory. This study was limited to a one-time, approximately one- hour session in which information on eight product categories 128 was collected. Brand loyalty over pap; purchasing was needed. Since longitudinal daily purchase data was unavailable, subjects had to remember what brands, if any, were purchased how many times in the prior months. Having subjects list and indicate a count of each brand purchased in each category over a particular time period would have been too unreliable given memory constraints and not standardized because of varying individual purchase cycles. Basing brand loyalty on brand purchase sequence would also be inaccurate in this situation because this would complicate the task by requiring individuals to recall, in addition to brand names and purchase amounts, the order in which the brands had been purchased. The reliability test-retest method of measuring brand loyalty stability across two time intervals was not possible without a longitudinal design. Instead, a constant sum scaling technique was used where each subject wrote in her two most often purchased brands and then divided the last ten purchases of the product category among the two brands plus an "all other brands" category. Since previous studies found the majority of consumers to purchase one or two favorite brands, having subjects list only two brands simplified the task and supplied sufficient data for brand loyalty analysis. The exact instructions read as follows: In the left column below list your most preferred or usually purchased brand first, then list your second most preferred or bought brand in each of the product categories. For the right-hand column, think back over your last ten purchases. Estimate the number of times out of ten that you purchased each of the brands you listed or "other brands." 129 These purchases should always add to ten for each product category. In the event that you have not purchased a product ten times, the total should correspond to the amount you have purchased. For each category, then, subjects who were users of the product were classified into one of four categories: (1) loyal to the test brand, (2) loyal to a brand other than the test brand, (3) not loyal but a purchaser of the test brand, or (4) not loyal and also not a purchaser of the test brand. This coding scheme allowed dichotomization of subjects in two ways for analysis purposes: as brand loyal or a brand switcher, and as a user of the test brand versus a non- regular test brand user. Loyalty for either the test brand or another brand (the first two categories) was defined in this case as eight out of ten purchases for one brand or 803 of a category’s purchasing. If a subject had purchased the product less than ten times, she was told to sum to the total number of pur— chases she had made, and these fractions were converted to percentages for classification purposes. If one brand was purchased three times out of four category purchases, this 75% was rounded to 80* and coded as loyal. Brand switchers were those people who divided the ten purchase sum among several brands, all of which were bought less than eight times. A threshold level of three was applied so that anyone purchasing the product less than three times was not included in the analysis. The concept of a threshold was also imple- mented by Cunningham (1956) so as not to artificially inflate 130 or deflate loyalty over the sample. This practice eliminates problems associated with coding the one—time purchase of a category where repeat purchase cannot be ascertained, as well as one out of two situations where brand loyalty has not been strongly established. The first instance would inflate loyalty while the second instance might deflate loyalty. The use of more specific coding categories such as dual loyalists was considered. However, given the small number of cases under analysis, splitting groups within a product cate- gory more narrowly would have proven meaningless; cell sizes would analytically be too small. The market share concept, where the proportion of purchases to the brand most. frequently purchased serves as a continuous loyalty scale ranging from 0% to 100%, was also considered and rejected. Transforming a ten purchase sequence into a 100 percentage point distribution is not warranted even assuming subjects could accurately recall purchase behavior. The constant sum scale serves better to reveal generalized brand purchasing tendencies given top-of—mind recall, than precise brand share. For this reason, the data was collapsed into broader categories for analysis. In essence, this brand share concept was initially used; by percentaging purchases for each brand before collapsing to standardize unequal category purchase amounts. For analysis purposes, to compare brand loyal persons to brand switchers (or those with low loyalty) would ultimately suggest dividing a continuous scale to distinguish between the two groups. The data was collapsed initially into broader categories for these reasons. 131 The brand loyalty question appeared early in the questionnaire after the general product usage questions so that aided brand cues would not bias the brand listings for past purchase. Coupon usage and deal proneness Because this experiment focused on coupon placement in media advertising, and because coupon redemption represents the primary dealing activity of consumers, coupon usage was the focus of the dealing questions. First, in a manner similar to that used by Strang and Gardner (1983) in their couponing study, subjects were asked to report about how many coupons they redeemed per month. Given that purchase cycles for toiletry products and health and beauty aids are longer than for grocery items, redemption per month, rather than Nielsen’s per week indicator, was thought to be sufficiently sensitive and discriminating. Due to a lack of data on the generality of deal prone— ness across product categories, the experimental subjects were also asked to specify if they had ever redeemed a coupon in each of the test product categories and to estimate how many of their last ten purchases were bought on deal or using coupons. From this ten—point scale a percentage of how much of the product was bought on deal was calculated. Except for the origination of the purchase information, this percentage is the same as Montgomery’s measure of deal proneness. It is similar to Webster’s DPI except for the omission of the adjustment for the market’s opportunity to deal. If a 132 subject had not purchased within a product category ten times, the percentage was calculated based on how many units she had purchased. By comparing the number of coupons redeemed per month with the coupon redemption for individual product categories, it was evident that dealing activity is product category specific. Some women reported a monthly redemption figure but also reported no redemption in any of the test product categories. These persons obviously redeem for grocery items or products not tested and must be considered deal prone to some degree. However, they are not deal prone for individual categories. Therefore, a general classification was desirable according to monthly redemption, as well as a product-by—product specific categorization. In this case, those women redeeming less than four coupons per month were non-deal prone and those redeeming four or more monthly were considered deal prone. An overall deal proneness index for only the test products was calculated by summing the number of deal purchases over all categories and taking that number as a percentage of all purchases on which the scale was based. Pretegting of thgyProcedure pad Ingtrument To pretest the procedure, the experimental disguise, and the self-administered questionnaire, an experimental session was conducted with several females in the target age group. This session was followed by an in-depth diagnostic focus 133 group interview to detect potential problems. At no point were the subjects told the true experimental purpose. Based on the feedback from this group, the introduction to the experiment which explained the disguised task involved was expanded in order to better explain the campaign concept to subjects without any advertising background. An informal comparison of the test and control adverti- sements in terms of creative appeal uncovered no major differences among them. One test advertisement was replaced because of comments on how "boring" the ad was compared to the others. Slides were originally timed for eight second exposure. Subjects, when asked, thought this was enough time but agreed that ten seconds might have been less rushed in some cases. Therefore, ten second exposures were used in the final experiment. Several minor changes in the questionnaire layout were instituted to facilitate understanding of the questions and make it appear less intimidating. For instance, space for open—ended descriptive responses was visually increased by widening the space between lines. Although the instructions were clear in instructing subjects to write down any descrip— tion of the ad they could remember, along with any thoughts they remembered having had while viewing the advertisement, the pretest showed that after the first few ads, subjects forgot to record any response for the second part. There- fore, for each test ad, the lines for the descriptive recall were subdivided into two parts titled "things" in the ad and "response" to the ad to act as a reminder prompt throughout. 134 Originally the constant sum scale which asked subjects to write down brands they had purchased and to divide ten purchases among them, allowed entry of four brands. Since subjects rarely wrote in four alternatives, the number of brands to be written in by the subject was reduced from four to two and an "all other brands" category added in its place. In addition, the question on product usage was changed by narrowing purchase frequency intervals to increase measurement sensitivity since most of the test products were purchased quite frequently. A "don’t purchase" option was added to several questions for those who weren’t users of a product category. No one seemed surprised or curious about the questions asked and how they related to the campaign disguise. For instance, the coupon questions, which were purposely under— played, did not arouse suspicion or comment. The group seemed to easily spot the campaigns, although everyone admitted that the campaign sequences were not immediately evident because many ads seemed to repeat. At least two of the three campaigns were unknown to each subject and the women seemed to enjoy the selected campaigns. The general response was enthusiastic and all the subjects thought the session was "fun." Data Analysis Procedures The primary statistical techniques planned for the analysis of the final data set, were chi-square, analysis of variance, and crosstabulations. For the nominal measures of 135 aided and unaided recall and coupon mention, two-way and three—way chi-square analyses were conducted to compare differences across repetition levels, between advertising and sales promotion, and across repetition while controlling for the promotional condition (advertising versus sales promotion). Coupon mention was also analyzed across exposure frequency for deal prone and non—deal prone subjects within the sample. For all other ratio-level measures--a cognitive response index value, brand evaluation, value and purchase intent--two and three-way analysis of variance techniques were used, both with and without the covariates of deal proneness, brand loyalty, and purchase frequency for the product category. Due to the diverse nature of the selected product categories, brands, and advertising executions, the eight product categories were analyzed individually rather than in aggregate. Also, differences in brand loyalty patterns, deal proneness and purchase frequency warranted separate evaluation. The results of these eight analyses follow in Chapter IV. Limitations Design Several limitations if this design deserve discussion. In order to control for exposure, subjects viewed the test advertisements in a laboratory environment and saw the print ads via slide presentation rather than in a portfolio or magazine. To control exposure frequency and exposure time 136 per ad, this laboratory slide presentation was most practical as discussed earlier. Also, the disguise and the abbreviated time period under which the experiment took place, helped to alleviate any reactance due to the experiment itself. This technique has been used by others in past advertising repeti— tion studies. Although ten minutes is a short time period in which to view 50 ads in close succession and in which to discriminate between experimental conditions related to time, realistically consumers spend at least ten minutes in one sitting reading one or more magazines in which advertisements are very densely spaced. In fact, the average subject for this study indicated that she spent 23 minutes daily with magazines. Additionally, in similar studies done in short time spans, researchers have found significant effects to occur between exposure conditions (Hockey, 1978). Variable Operationalizations Since couponing is the major price promotion technique, comprising almost 70% of the dollars spent on promotion, coupons were selected to represent the price promotion variable rather than rebates, premiums, or on—pack price offs. So justified, it should be noted that response to coupons probably is different than for other sales promotion techniques. Print ads were used rather than television commercials. As print tends to be a higher involvement medium than televi- sion, intermedia generalization is difficult. However, since 137 price promotions, especially coupons, are most commonly delivered via newspapers, free standing inserts, Sunday supplements, and magazines, print was felt to be the best choice in this experiment. In terms of time constraints, print seemed more expedient, especially when testing several product categories. Repetition of 30—second television commercials within any relatively short time span would be practically precluded. All advertisements were currently running in the media at the time of the study and it is to be expected that some prior exposure to the ad and brand may have occurred. In fact, in some cases, the print ads were part of larger cam- paigns involving television. However, advertisers are rarely in a position to test advertising in an environment free of brand awareness and knowledge; therefore, this design is realistic. Securing ads for new products or new not-yet- aired ads for established brands was beyond the time and cost constraints for this study. For fictitious brand dummy ads more exposure and time would be necessary just to introduce the brand and it would be unrealistic to expect attitudes to develop in such a brief time span. Furthermore, much of the worry behind couponing is associated with its continued use and effect on established brands since promotions are almost always used in new product introductions initially and consumers recognize and understand this fact. Scott (1976) and Aaker (1972) have suggested that magni- tude of the deal is inversely related to brand image and that to properly test for promotion effects, the coupon face value 138 should vary along a continuum. For this design, coupon face values were left as found in the media, therefore, represen- ting what the industry considers acceptable price-off values. These, then, were manipulated merely as present or absent to establish first that effect gggg, in fact, vary. According to Nielsen’s consumer study (1985), one third of those who use coupons are not influenced by the magnitude of the deal at all. CHAPTER IV ANALYSIS OF EXPERIMENTAL RESULTS In this chapter, the general sample characteristics are described first. The final data recoding, along with the computation of the cognitive response valence indexes, are explained and the data analysis procedures are outlined as they will be presented in the individual product category analyses. The significant findings from each of the eight product category analyses are then described and discussed for each criterion measure in sequence. The categories analyzed, in order, are: (1) hair remover, (2) nail polish remover, (3) mascara, (4) headache remedies, (5) powdered drink mixes, (6) sanitary napkins, (7) deodorant, and (8) face and body soap. Sample Characteristics Demographics Table 4.1 shows the demographic composition of this sample across several variables. Each of these demographic characteristics will be briefly discussed. The women in the sample ranged in age from 18 to 44 years, with an average age of 20. Seventy three percent of the participants were between 19 and 21 years of age. Only one woman fell outside the target age range of 18 to 34. This 44 year old woman, a student, was included because she 139 140 indicated regular readership of the print vehicles from which the test ads were taken. There were no significant deviations among the average ages of the ten cell conditions. The overwhelming majority, 95%, were single, with 90% living alone or with roommates. Ninety seven percent were enrolled as students at one of the local colleges or universities in the community. Table 4.1 reveals that the majority (79%) were sophomores, juniors and seniors at the time. They resided in university on- campus residence halls, apartments, and off-campus houses primarily. TABLE 4.1 SAMPLE DEMOGRAPHIC CHARACTERISTICS Demographic Variable Classification Percentage Age 18 years 1% 19 25 20 29 21 19 22 15 23 and over 12 Sample Living Situation Living with roommates 80% Living alone 10 Living with parents 7 Living with spouse 3 Student Status Graduate students 6% Seniors 25 Juniors 35 Sophomores 29 Freshmen l Non-student 3 Student Residency Residence halls 35% Apartment 34 House 24 Fraternity 7 141 Thirty nine percent of the sample stated they were not primary purchasers of grocery products, which is not surprising given the number living in residence halls where meals are prepared, or with roommates who might share the shopping task. Grocery products were specifically avoided as test product categories for just this reason. The only food- type product, a powdered drink mix, was included because it would most likely be purchased regardless of whether or not the individual grocery shopped or regularly prepared her own meals. For personal products, 94% claimed full purchase responsibility. Seven of the the eight test product categories would be classified as personal in nature. Media Usage Over all subjects, the average number of television viewing hours per day was about two hours, with 45% watching less than this and 24% watching more than this (3+ hours per day). On a typical day, these women spent about 20 minutes reading magazines. Only 16% said they spent a half hour or more per day with magazines. Everyone in the sample claimed to regularly read at least one of the print vehicles from which test ads were taken which confirms that everyone was in the appropriate target market for these advertised test brands. Product Usage The top two rows in Table 4.2 show the percentage of sample subjects in each test product category who claimed to 142 purchase this product as frequently as once every month and the percentage who had purchased at least once in the six months prior to the experiment. As shown, there is some variance among categories. Soap and deodorant are the most frequently purchased products and nearly everyone in the sample had purchased these two products relatively recently. The least purchased products were hair remover and nail polish remover, with purchase cycles of longer than three months. In all but one category, a large majority, over 50%, had purchased within a six month time period. The percentage purchasing in the prior months may be slightly lower-than- average for these students due to the fact that many had just returned to school from living at home during summer break where they had relied on other family members to actually purchase some of these products such as soap. Brand Loyalty Table 4.2 also shows the percentage who are brand loyal in each product category. It is apparent that some categories have higher brand loyalty than others. Nail polish remover and deodorant ranked highest in brand loyalty. Most likely, this is due to the few number of national brands in the polish remover category, while the personal, higher- risk nature of a personal hygiene product like deodorant, along with the high differentiation among national deodorant brands, could account for the deodorant loyalty. Mascara, headache remedies, and sanitary napkins all have over 40% who 143 TABLE 4.2 PURCHASING PATTERNS, BRAND LOYALTY AND COUPON USAGE 0F SAMPLE % s "U H o x H E a o m o m > m m z 0 :H X H E H a) ‘H >» c: w o u m a E H m Dd ‘34 0) H U (U H > m m x u o H r—1 0 U 'U C: 'H "U D- ‘H -.—4 E to m ‘H c: o a: (V1 C0 (1) (0 Cl.) H (U 0.) O m Z Dd >2 3: Q m a to Purchase once per month 2% 4% 5% 11% 22% 26% 45% 61% Purchased in last six months 16 62 84 76 52 73 97 99 Brand loyal 16 56 46 46 28 44 56 40 Redeem coupons in category 10 29 22 37 38 42 54 57 Mean number of last ten purchases bought .41 .86 .59 1.62 1.51 2.12 2.39 2.56 on deal. are loyal. The lowest loyalty falls to hair remover which coincides with the purchase characteristics of this product. In terms of the patterns among individuals, brand loyalty is a selective behavior rather than a general behavior trait. By aggregating over all eight test product categories it was possible to see in how many of the eight categories an individual was brand loyal. Table 4.3 shows that no one in the sample was brand loyal in all eight product categories, although 25% of the women were loyal to at least five of the eight categories. Very few women in the sample, under 5%, were not loyal to any of the eight categories. 144 TABLE 4.3 GENERAL BRAND LOYAL AND DEALING BEHAVIOR Number of Percentage Percentage Categories Brand Loyal Who Use Coupons 0 4% 27% 1 10 5 2 17 ll 3 25 15 4 19 15 5 17 12 6 7 10 7 l 4 8 0 l 100% 100% Deal Proneness To get a feel for the dealing tendencies of this sample, everyone was asked to report about how many coupons they redeemed per month over all product categories. The median number of coupons redeemed was 3.0 with a range of zero to 50 per month. This large range accounts for the positive skew in the mean which was 6.2. Table 4.4 reveals that the magni- tude of dealing is evenly spread throughout the sample. Twenty two percent claimed no coupon dealing activity, while 78% did. The latest Nielsen report of nationwide coupon statistics revealed that 79% of the consumers redeem coupons (Nielsen 1985). This percentage includes redemption in all product categories, including grocery product coupon redemption. The high correlation between this sample and the population on this usage statistic helps to verify the representativeness of this sample as a test group. 145 TABLE 4.4 AVERAGE COUPON REDEMPTION PER MONTH Number of Coupons/Mo. Percentage 0 22% 1-3 28 4-9 23 10—19 18 20—50 9 100% Coupon usage, like brand loyalty, appears to be category specific, suggesting that people may be deal prone on a selective basis. Table 4.2 shows redemption patterns over the eight test product categories. Deal usage ranges from 10% in the least frequently purchased category, hair remover, to 57% in the most frequently purchased category, soap. There is a high positive relationship between coupon redemption and purchase frequency. Table 4.3 shows the extent of coupon usage for an individual over all eight test product categories. Twenty seven percent claim to have never redeemed a coupon for any of these products. There appears to be a slightly larger tendency for a person who uses coupons to deal in all categories. Five percent deal in seven or eight of the categories, whereas only one percent were brand loyal in seven of the categories and none were brand loyal to eight. According to these numbers, 73% of the sample claim to use coupons in at least one of the selected categories. Again, compared to a national redemption statistic of 79% over all 146 product purchasing, this 73% for a few categories is right on target. From this discussion it appears that individual category differences do exist. For example, soap is a low loyalty, high dealing category relative to the nail polish remover category, which would appear to be highly loyal but less prone to dealing. It is probably reasonable to assume that these figures are reflective of the opportunity to deal in the category. The characteristics of each test category will be discussed in more detail with respect to the statistical analyses in the experimental findings. Data Recoding and Analysis Procedures Chapter III described the recoding of the basic measures and initial coding schemes. For analysis purposes, many of these variables were recoded into broader categories for chi square analysis. With 393 cases, and only 30 to 40 subjects per cell, the data would not allow much more than two-way divisions within a cell. In any case where less than five subjects fell into a cell and it was not otherwise possible to collapse the categories further, Yates correction factor was applied to correct for the small cell frequencies. Two examples of this collapsing are especially relevant. Advertising recall was collapsed into two categories: (1) no or non—qualified recall and (2) qualified recall. The brand loyalty covariate was also treated as a dichotomy: (1) brand loyal to either the test brand or another brand, or (2) not brand loyal. 147 The cognitive response data was manipulated in several ways. In the original data, many women wrote several thoughts and each was coded individually. First, to obtain an overall net valence index, the number of positive thoughts, including support arguments, product affirmations, and positive ad evaluations, were totaled for each individual. From this, the number of negative responses, including counterarguments, disaffirmations, and negative ad evaluations, was subtracted. This computation resulted in a net valence score ranging from -3 to +3. To compute a product-related valence index, only support arguments and affirmations were included in the positive count and only counterarguments and disaffirmations were counted as negative. The net valence was calculated in the same way. Ad-related cognitive response indicators only included positive and negative ad evaluations in the computation of the index. This indicator ranged from -1 to +1 since few individuals mentioned much more than a general statement of like or dislike for the ad—as—a—whole or a particular element. These net valence indexes, then, were treated as interval scales and examined with analyses of variance. Individual Category Analyses Preliminary data analyses which aggregated the eight test product categories across all measures did not show significance on these measures due to the extreme 148 differences, in some cases, among the category characteristics and advertising executions. Therefore, the analysis presented here is not aggregated. Each category is treated as a separate test of the hypotheses listed in Chapter II and the individual category analyses shown for each measure. Hair Remover Hair remover is an infrequently purchased product with the majority of those purchasing in the category (78%) purchasing it less often than every three months. There is also low loyalty and coupon usage in this category. In fact, hair remover ranked the lowest in all three respects: purchase frequency, brand loyalty and dealing. Table 4.5 shows the individual cell means across repetition for both advertising and sales promotion, as well as the "total" grand means for these conditions. The last column shows the significance values. For brand and advertising-related recall there were no significant main effects for either repetition or promotion using chi square analyses. Analyses of variance performed on the value ratings, brand evaluations, and purchase intentions are also shown in Table 4.5. The significance values off to the right are, in order, for the repetition main effect, the promotion main effect, and the interaction term. As shown, the "value" criterion variable showed a signi- ficant main effect for promotion at the .03 probability level. It appears that value becomes more negative over 149 TABLE 4.5: HAIR REMOVER EFFECTS Exposures l 2 3 Total Prob. Brand Recall Total 69.2 70.8 63.0 67.5 .54 Advertising 71.8 78.8 62.5 70.5 .31 Sales Promotion 66.7 64.1 63.4 64.7 .95 Adv. vs S.P. .42 Related Recall Total 82.1 91.7 81.5 84.8 .15 Advertising 89.7 90.7 82.5 87.5 .49 Sales Promotion 74.4 92.3 80.5 82.4 .11 Adv. vs S.P. .37 Value Covariates: Purchase Frequency .46 Brand Loyalty .59 Coupon Experience .44 Total 5.83 5.71 5.90 5.81 Advertising 6.15 4.62 5.17 5.32 .85 Sales Promotion 5.00 7.00 6.89 6.56 .03* Interaction .14 Evaluation Covariates: Purchase Frequency .04% Brand Loyalty .89 Coupon Experience .55 Total 2.33 2.25 2.67 2.41 Advertising 2.15 2.38 3.08 2.53 .32 Sales Promotion 2.80 2.09 2.11 2.24 .06 Interaction .07 Purchase Intention Covariates: Purchase Frequency .07 Brand Loyalty .80 Coupon Experience .88 Total 5.28 5.38 5.10 5.25 Advertising 5.62 4.46 4.58 4.89 .98 Sales Promotion 4.40 6.45 5.87 5.80 .10 Interaction .40 150 repetition with advertising and more positive over repetition with sales promotion, although at the three exposure level, the trend slightly reverses itself. There is a significant covariate, purchase frequency, on brand evaluation which, once controlled, almost makes the promotion main effect significant (.06). Again, the trend shows increased use of sales promotion increases brand evaluation. (Lower means indicate more positive ratings for brand evaluation.) Purchase intent shows no significant effects for hair remover, although sales promotion elicits more positive intentions at the higher exposure levels as compared to advertising. Table 4.6 shows the results of the cognitive response data. Significant promotion main effects are revealed for the overall net valence indicator and for the ad-related response valence. In both cases, sales promotion produces significantly more favorable comments than advertising at all repetition levels. All indexes are negative for this advertisement but sales promotion is much less negative on the first two exposures than advertising. In the coding of this particular advertisement, a great many subjects thought the focus on the mid-portion of a female body distasteful and sexist. However, from these results it appears as though the coupon in the advertisement made the ad more favorable, perhaps by distracting from the negative aspects of the ad. The difference in product related responses for advertising and sales promotion were 151 not quite significant at the .05 probability level (.07) but the sales promotion condition experienced less counterargumentation than advertising. TABLE 4.6 HAIR REMOVER COGNITIVE RESPONSE Exposures l 2 3 Total Prob. Overall Total —.33 -.21 —.38 —.30 Advertising -.54 -.46 -.50 -.50 .69 Sales Promotion .20 .09 -.22 .00 .003* Interaction .55 Product Total —.15 —.04 -.06 —.09 Advertising -.23 —.03 -.18 —.15 .39 Sales Promotion —.08 -.05 .05 -.03 .07 Interaction .33 Advertising Total —.26 -.39 —.31 —.32 Advertising -.41 —.45 -.33 -.39 .31 Sales Promotion —.10 -.33 —.29 -.24 .04* Interaction .28 There does seem to be truth in this category to the hypothesis that deal prone persons would notice a coupon in the advertisement to a greater degree than non—deal prone persons. The deal prone noticed the coupon four times more than those who do not use coupons. This was significant at the .019 level as Table 4.7 shows. 152 TABLE 4.7 HAIR REMOVER COUPON MENTION BY DEAL PRONENESS Non—Deal Deal Prone Prone Total Prob. Percentage Recall 12.6 50.0 15.1 .019* Nail Polish Remover Nail polish remover is also an infrequently purchased product with almost 56% of those purchasing claiming to purchase less than every three months. This was one of the most brand loyal categories, perhaps because there are so few major brands. This category is characterized by low coupon redemption. In Table 4.8, the means and significance values for the main effects are shown. Significant repetition effects are found for both brand and advertising—related recall across repetition for both advertising and sales promotion. This is a result of the dramatic increase in recall found between the first and second exposures. Again, a significant promotion main effect is found on ' Again, sales promotion increased the value of the "value.' product as compared to advertising. Both purchase frequency and brand loyalty were significant covariates on this variable. Neither brand evaluation or purchase intent showed any significant results. 153 TABLE 4.8: NAIL POLISH REMOVER EFFECTS Exposures l 2 3 Total Prob. Brand Recall Total 77.4 91.8 88.8 85.7 .02: Advertising 72.7 89.7 87.8 83.1 .07 Sales Promotion 82.5 94.1 89.7 88.5 .28 Adv. vs S.P. .32 Related Recall Total 45.2 80.8 80.0 67.9 .0008 Advertising 43.2 74.4 78.0 64.5 .001* Sales Promotion 47.5 88.2 82.1 71.7 .OOOt Adv. vs S.P. .30 Value Covariates: Purchase Frequency .014 Brand Loyalty .03! Coupon Experience .15 Total 7.89 8.39 7.89 8.04 Advertising 7.53 7.97 7.73 7.74 .10 Sales Promotion 8.28 8.85 8.05 8.38 .031 Interaction .81 Evaluation Covariates: Purchase Frequency .56 Brand Loyalty .06 Coupon Experience .59 Total 1.68 1.53 1.49 1.57 Advertising 1.72 1.61 1.54 1.62 .12 Sales Promotion 2.80 2.09 2.11 2.24 .20 Interaction .96 Purchase Intention Covariates: Purchase Frequency .000% Brand Loyalty .OOOt Coupon Experience .01* Total 8.26 8.47 7.95 8.22 Advertising 8.07 8.13 7.68 7.96 .17 Sales Promotion 8.46 8.85 8.23 8.50 .18 Interaction .84 154 The cognitive response data showed no differences in the overall net valence and the product valence; however the ad- related valences were significant for both repetition effects and promotional conditions. In fact, the interaction term was nearly significant on this variable. Advertising response was not as volatile as sales promotion from one exposure level to the next. Response for both is shaped in an inverted ”V" with negative response at both one and three exposures and positive response with just two repetitions. However, advertising was less negative in the extremes and less positive on the second exposure, which would account for the interaction between repetition and promotion (Table 4.9). TABLE 4.9 NAIL POLISH REMOVER COGNITIVE RESPONSE Exposures l 2 3 Total Prob. Overall Total -s01 s15 003 005 Advertising -.02 .08 .17 .07 .35 Sales Promotion .00 .24 -.13 .03 .63 Interaction .14 Product Total .06 .ll .20 .12 Advertising .05 .03 .22 .10 .37 Sales Promotion .08 .21 .18 .15 .53 Interaction .56 Advertising Total -.07 -.04 -.16 -.07 Advertising -.05 -.03 -.05 -.02 .003t Sales Promotion -.10 -.06 -.28 -.12 .002* Interaction .07 155 Table 4.10 reveals that deal prone persons again noticed coupons more than non-deal prone people; however, this difference was not significant for nail polish remover. This could be due to the fact that this particular ad mentioned the price—off in the headline for the sales promotion condition: "15 off the remover that makes nails stronger every time you use it." Therefore, even non—deal-prone subjects would likely be drawn to the coupon. TABLE 4.10 NAIL POLISH REMOVER COUPON MENTION BY DEAL PRONENESS Non-Deal Deal Prone Prone Total Prob. Percentage Recall 48.1 66.7 54.0 .099 Mascara Mascara has about a three month purchase interval for the majority of the sample (40%). It has relatively high brand loyalty with almost half the sample claiming to be brand loyal (46%). This loyalty may possibly be overstated just a bit if an individual bought several varieties under one brand name but didn’t indicate this. For instance, some subjects wrote in "Cover Girl" while others wrote in "Long-n- Lush" or "Marathon," both Cover Girl products. Coupon redemption is low for this category as it is for hair remover and polish remover. The overall means for brand recall just barely show significance over the three repetition levels as shown in 156 Table 4.11. Yet, neither advertising or sales promotion show significance individually. This could be due to the larger sample size of the two promotion conditions aggregated. As expected, recall increases with each additional exposure. The incremental increase between the first two exposures exceeds that of the second to the third exposure, demonstrating the diminishing returns of the learning curve. Advertising shows a significant repetition effect for advertising-related recall due to a 26% increase between the first and second exposures, although sales promotion does not. In this case, sales promotion has higher recall levels at both the second and third exposure level. None of the other criterion measures revealed any significant patterns of response over promotion or repetition. Table 4.12 shows the results for the cognitive response data. ——7 157 TABLE 4.11: MASCARA EFFECTS Exposures 1 2 3 Total Prob. Brand Recall Total 84.6 92.3 95.8 90.8 .05* Advertising 84.6 94.1 97.4 92.0 .10 Sales Promotion 84.6 90.9 93.9 89.7 .41 Adv. vs S.P. .71 Related Recall Total 74.4 92.3 93.1 86.4 .001* Advertising 71.8 97.1 97.4 88.4 .000* Sales Promotion 76.9 88.6 87.9 84.5 .28 Adv. vs S.P. .50 Value Covariates: Purchase Frequency .35 Brand Loyalty .16 Coupon Experience .06 Total 5.31 5.45 5.45 5.40 Advertising 5.42 5.40 5.46 5.43 .91 Sales Promotion 5.22 5.49 5.43 5.38 .83 Interaction .97 Evaluation Covariates: Purchase Frequency .002* Brand Loyalty .83 Coupon Experience .03* Total 2.91 2.82 2.77 2.83 Advertising 2.91 2.80 2.86 2.86 .69 Sales Promotion 2.92 2.83 2.67 2.81 .85 Interaction .66 Purchase Intention Covariates: Purchase Frequency .15 Brand Loyalty .02* Coupon Experience .02* Total 3.50 3.77 3.74 3.67 Advertising 3.82 4.03 3.49 3.77 .75 Sales Promotion 3.22 3.59 4.03 3.58 .54 Interaction .64 158 For this category, deal prone persons noticed the coupon in the sales promotion condition more than twice as much as non-deal prone persons as Table 4.13 shows. This difference was significant at the .032 probability level. TABLE 4.12 MASCARA COGNITIVE RESPONSE Exposures 1 2 3 Total Prob. Overall Total —.33 —.18 -.28 —.26 Advertising -.09 —.20 -.26 -.18 .61 Sales Promotion —.54 —.17 —.30 -.33 .23 Interaction .27 Product Total -.18 —.l4 —.l8 —.17 Advertising .00 —.12 —.21 —.11 .90 Sales Promotion -.36 -.16 —.15 -.22 .24 Interaction .23 Advertising Total —.06 -.01 -.07 —.05 Advertising —.03 .03 -.03 -.01 .57 Sales Promotion -.10 -.05 -.12 -.09 .14 Interaction .99 TABLE 4.13 MASCARA COUPON MENTION BY DEAL PRONENESS Non-Deal Deal Prone Prone Total Prob. Percentage Recall 16.1 39.1 20.7 .032* —— . . . .11... 159 Headache Remedies About 30% of the sample purchase headache remedies every three months or so, while 35% purchase this product more frequently and 35% purchase less frequently. About half the sample claim to be brand loyal in this category (46%) and another large minority, 38%, used coupons. It is interesting that headache remedy loyalty is equal to and not higher than mascara loyalty given the product nature. Table 4.14 shows that both brand and advertising recall measures were significant across repetition levels for both sales promotion and advertising conditions. There was no promotion effect. The recall percentages are higher across the board for advertising than they are for sales promotion in this product category. Value ratings, brand evaluations, and purchase intentions were not significant across either of the main effects. The differences in means are slight and not significant but it appears that advertising creates higher "value" in the consumers mind for the brand, as well as improving brand evaluations (lower brand evaluation means indicate more favorable impressions). 160 TABLE 4.14: HEADACHE REMEDY EFFECTS Exposures l 2 3 Total Prob. Brand Recall Total 64.9 83.3 88.2 78.9 .001* Advertising 70.7 87.2 90.9 82.3 .05! Sales Promotion 57.6 79.5 86.0 75.7 .01* Adv. vs S.P. .29 Related Recall Total 59.5 80.8 92.1 77.6 .000* Advertising 63.4 87.2 97.0 81.4 .001* Sales Promotion 54.5 74.4 88.4 73.9 .004# Adv. vs S.P. .23 Value Covariates: Purchase Frequency .59 Brand Loyalty .57 Coupon Experience .71 Total 6.85 6.38 6.25 6.48 Advertising 7.00 6.72 6.50 6.74 .39 Sales Promotion 6.71 5.90 6.06 6.22 .09 Interaction .81 Evaluation Covariates: Purchase Frequency .37 Brand Loyalty .43 Coupon Experience .19 Total 2.42 2.44 2.43 2.43 Advertising 2.33 2.41 2.42 2.39 .88 Sales Promotion 2.50 2.48 2.44 2.47 .34 Interaction .94 Purchase Intention Covariates: Purchase Frequency .40 Brand Loyalty .56 Coupon Experience .83 Total 5.19 4.84 5.55 5.21 Advertising 4.83 5.17 5.96 5.31 .32 Sales Promotion 5.54 4.38 5.25 5.10 .50 Interaction .28 161 Analyses of variance on the cognitive response data also showed nothing significant or very consistent as reported in Table 4.15. TABLE 4.15 HEADACHE REMEDY COGNITIVE RESPONSE Exposures 1 2 3 Total Prob. Overall Total .29 .16 .13 .19 Advertising .33 .28 .04 .22 .51 Sales Promotion .25 .00 .19 .16 .61 Interaction .36 Product Total .08 .00 .12 .07 Advertising .07 .13 .12 .ll .52 Sales Promotion .09 -.13 .12 .03 .55 Interaction .38 Advertising Total .01 .06 -.05 .01 Advertising .02 .05 -.03 .02 .11 Sales Promotion .00 .08 -.07 .00 .79 Interaction .82 In addition, only a slight difference appeared in the recall of coupons for deal prone people as compared to non— deal prone people. This was not significant. TABLE 4.16 HEADACHE REMEDY COUPON MENTION BY DEAL PRONENESS Non-Deal Deal Prone Prone Total Prob. Percentage Recall 12.5 18.6 14.8 .535 162 Powdered Drink Mixes A bimodal purchase frequency distribution is apparent for powdered drink mixes. Just over one third (34%) purchase this product category every month while this is balanced by a large group (37%) who purchase it less than every three months. Few are loyal——this is the second lowest ranked category in this respect--and coupon redemption is average compared to the other test product categories. There is a little ambiguity as to exactly how this product category is defined and the lower brand loyalty figure may be a reflection of this. Some women in the sample included hot chocolate drink mixes and herbal teas in this category while others saw it strictly as a lemonade, punch, and Koolaid category. Although this definition problem in categorizing loyalty has not been discussed, it is an issue that is consistently a problem in the analysis of loyalty in scanner data. Because it is the respondents themselves who have individually defined the category as they perceive it, and not an outside arbitrator deciding what will constitute the category competition, this data may be more realistic than that of past studies. Of all the measured variables, reported in Table 4.17, only brand recall showed any significance for this drink mix. Both advertising and sales promotion displayed repetition effects. The primary difference again is the jump between the first and second exposures. The ad recall percentages showed more gradual increases across repetition, and these were not significant. 163 TABLE 4.17: DRINK MIX EFFECTS Exposures l 2 3 Total Prob. Brand Recall Total 57.3 83.8 83.6 74.5 .000* Advertising 59.0 84.6 78 8 73.9 .03* Sales Promotion 55.8 82.9 87.5 75.0 .001* Adv. vs S.P. .96 Related Recall Total 75.6 82.5 86.3 81.3 .22 Advertising 76.9 79.5 81.8 79.3 .88 Sales Promotion 74.4 85.4 90.0 83.1 .15 Adv. vs S.P. .57 Value Covariates: Purchase Frequency .01! Brand Loyalty .02* Coupon Experience .46 Total 5.85 5.92 5.53 5.78 Advertising 5.78 5.65 5.25 5.57 .66 Sales Promotion 5.91 6.21 5.78 5.97 .37 Interaction .86 Evaluation Covariates: Purchase Frequency .02* Brand Loyalty .003# Coupon Experience .30 Total 2.34 2.10 2.44 2.29 Advertising 2.56 2.10 3.44 2.35 .27 Sales Promotion 2.17 2.11 2.44 2.23 .28 Interaction .64 Purchase Intention Covariates: Purchase Frequency .01* Brand Loyalty .000* Coupon Experience .31 Total 6.15 7.05 5.91 6.39 Advertising 6.00 7.15 5.50 6.28 .29 Sales Promotion 6.26 6.95 6.28 6.48 .48 Interaction .54 — .’_)_\ W,“ .. , .V , 164 Purchase frequency and brand loyalty were significant covariates on the value, evaluation and purchase intention measures. This may be in part due to the large variance in frequency and redemption distributions for this category. The overall valence index for cognitive response and the ad-related responses (Table 4.18) showed no signs of significance. However, the product—related responses did vary significantly between advertising and sales promotion. Sales promotion increased the positive product comments drastically for the second and third exposure levels over the advertising condition. This is similar to what was found for the hair remover category. TABLE 4.18 DRINK MIX COGNITIVE RESPONSE Exposures l 2 3 Total Prob. Overall Total -.21 —.08 —.23 -.17 Advertising -.26 —.21 -.36 -.27 .48 Sales Promotion —.16 .05 —.13 —.08 .10 Interaction .82 Product Total —.06 .11 .08 .04 Advertising —.10 .00 —.06 -.05 .23 Sales Promotion —.02 .22 .20 .13 .04* Interaction .68 Advertising Total —.13 —.18 -.22 —.17 Advertising —.l3 —.l8 —.l5 -.15 .58 Sales Promotion —.l4 -.17 -.28 —.19 .55 Interaction .68 165 However, the deal prone were not more likely to notice the coupon in this advertisement. In fact, this is one of two categories where the deal prone remembered the coupon less than the non-deal prone. The difference is very slight and not at all significant. TABLE 4.19 POWDERED DRINK MIX COUPON MENTION BY DEAL PRONENESS Non-Deal Deal Prone Prone Total Prob. Percentage Recall 28.8 27.3 28.2 1.000 Sanitary Napkins Sanitary napkins are purchased frequently by this sample. Thirty four percent buy every month and another 38% purchase bimonthly. Brand loyalty and coupon usage are both in the moderately high range with 44% and 42% of the sample falling into each category, respectively. For brand and advertising recall, the sales promotion condition showed significant main effects along repetition which was enough to carry an overall main effect on these variables. The recall in both the advertising and sales promotion conditions increased steadily over repetition, but on the third exposure, sales promotion, which had been lagging behind, jumped to the lead in recall (See Table 4.20). The coupon usage or deal proneness covariate in this category was significant for all three higher level attitudinal measures. TABLE 4.20: 166 SANITARY NAPKINS EFFECTS Brand Recall Total Advertising Sales Promotion Adv. vs S.P. Related Recall Total Advertising Sales Promotion Adv. vs S.P. Value Covariates: Purchase Frequency Brand Loyalty Coupon Experience Total Advertising Sales Promotion Interaction Evaluation Covariates: Purchase Frequency Brand Loyalty Coupon Experience Total Advertising Sales Promotion Interaction Purchase Intention Covariates: Purchase Frequency Brand Loyalty Coupon Experience Total Advertising Sales Promotion Interaction 36.3 33.3 66.3 66.7 6.87 7.09 6.62 2.26 2.27 2.24 5.79 5.82 5.76 Exposures 2 47. 46. 75.0 79.1 6.92 6.59 7.30 2.36 2.32 2.40 6.02 6.12 5.90 3 58.3 54.5 61.5 88.9 84.8 92.3 7.31 7.41 7.24 2.29 2.32 2.28 6.41 5.91 6.79 Total 47.0 47.0 47.1 76.3 73.0 79.3 7.02 6.98 7.06 2.31 2.30 2.31 6.05 5.96 6.15 Prob. .02* .04* 1.00 .004* .02! .33 .15 .002! .45 .18 .22 .001* .67 .71 .20 .22 .000* .62 .25 167 According to Table 4.21, the overall cognitive response index was not significant but advertising was consistently more positive than sales promotion. Ad-related response showed advertising to be more positive as well, although not significantly more favorable. A significant overall repetition effect for product-related response appeared, but due to the smaller sample sizes of the individual conditions, this significance did not appear for advertising or sales promotion individually. TABLE 4.21 SANITARY NAPKIN COGNITIVE RESPONSE Exposures 1 2 3 Total Prob. Overall Total .02 .08 .24 .10 Advertising .06 .15 .41 .18 .37 Sales Promotion —.03 .00 .10 .02 .21 Interaction .82 Product Total .05 -.04 .25 .08 Advertising .00 —.07 .30 .06 .05* Sales Promotion .10 .00 .21 .10 .74 Interaction .66 Advertising Total -.05 .08 .06 .03 Advertising .00 .15 .09 .08 .17 Sales Promotion —.10 .02 .03 —.02 .12 Interaction .93 This is the second product category where deal prone persons recalled the coupon less often than the non-deal prone subjects, perhaps due to the size of the coupon which 168 extended across the bottom of the ad. Again, the difference was small and insignificant. (See Table 4.22) TABLE 4.22 SANITARY NAPHIN COUPON MENTION BY DEAL PRONENESS Non-Deal Deal Prone Prone Total Prob. Percentage Recall 22.9 15.7 19.8 .456 Deodorant Deodorant is one of the most frequently purchased products of the eight product categories. Forty percent purchase deodorant approximately once a month, while another 38% purchase every two months. This product is in the highest loyalty bracket, tied with polish remover for the largest brand loyal segment of the eight categories tested. As was mentioned earlier, this loyalty might be due to the perceived risk accompanying the product. Deodorant is also one of the categories where coupon redemption is prevalent. Over half the sample deal in this category (54%) and the average woman bought 2.4 of her last 10 deodorant purchases with a coupon, which is 25% of her deodorant purchasing. In this category, Table 4.28 shows repetition to experience a significant main effect for sales promotion brand recall and for the overall brand recall means but shows advertising to fall just short of significance along this dimension. Sales promotion brand recall is higher (not significantly) than advertising brand recall and steadily 169 increases over exposure levels. The largest increase is between the one and two exposure conditions with increasing returns between two and three repetitions only at a diminishing rate. Advertising-related recall has not only a significant repetition main effect for both promotion conditions, but also a promotion main effect. Advertising elicits significantly higher advertising-related recall than sales promotion and jumps from 48% recall at one exposure to 83% after three exposures. Sales promotion has significantly less recall after one exposure but closes the gap after the third by jumping to 82%. The coupon may be distracting from the advertising recall. There were no significant results in the higher level hierarchy of effects criterion variables. The cognitive response data showed one significant effect. The overall product-related valence indexes were significant across repetition. This relationship is shown in Table 4.24 Product-related response was initially positive, then negative, and finally positive for sales promotion, whereas advertising was always hovered close to neutral. Interestingly, the advertising response became more negative over repetition for sales promotion and more positive over repetition for advertising, perhaps showing increased tedium with the coupon. Deal prone persons in this category did not recall the coupon significantly more than non—deal prone persons. However, they do remember them better as Table 4.25 shows. 170 TABLE 4.23: DEODORANT EFFECTS Exposures l 2 3 Total Prob. Brand Recall Total 44.6 64.6 73.0 60.9 .001* Advertising 42.4 60 5 68.3 58.1 .08 Sales Promotion 46.3 69.2 78.8 63.7 .01* Adv. vs S.P. .46 Related Recall Total 40.5 64.6 82.4 62.6 .000* Advertising 48.5 74.4 82.9 70.1 .004* Sales Promotion 34.1 53.8 81.8 54.9 .000* Adv. vs S.P. .03* Value Covariates: Purchase Frequency .63 Brand Loyalty .05t Coupon Experience .83 Total 6.43 6.24 6.67 6.43 Advertising 6.44 5.83 6.76 6.32 .55 Sales Promotion 6.42 6.71 6.54 6.55 .60 Interaction .32 Evaluation Covariates: Purchase Frequency .56 Brand Loyalty .13 Coupon Experience .72 Total 2.52 2.64 2.51 2.56 Advertising 2.52 2.75 2.58 2.63 .59 Sales Promotion 2.53 2.52 2.42 2.49 .35 Interaction .68 Purchase Intention Covariates: Purchase Frequency .48 Brand Loyalty .005* Coupon Experience .55 Total 5.65 4.96 5.74 5.43 Advertising 5.70 4.56 5.70 5.27 .31 Sales Promotion 5.61 5.42 5.79 5.59 .74 Interaction .63 171 TABLE 4.24 DEODORANT COGNITIVE RESPONSE Exposures 1 2 3 Total Prob. Overall Total .13 -.04 .26 .11 Advertising .00 -.03 .24 .07 .07 Sales Promotion .22 —.06 .29 .14 .49 Interaction .61 Product Total .11 -.02 .22 .10 Advertising .06 .05 .15 .09 .02* Sales Promotion .15 -.10 .30 .11 .74 Interaction .19 Advertising Total .01 .00 .03 .01 Advertising -.03 -.02 .07 .01 .92 Sales Promotion .05 .03 -.03 .02 .86 Interaction .36 TABLE 4.25 DEODORANT COUPON MENTION BY DEAL PRONENESS Non—Deal Deal Prone Prone Total Prob. Percentage Recall 7.3 17.2 12.4 .186 Face and Body Soap All but one percent of the sample had purchased soap in the six months prior to when this study took place and 61% purchase it every month. Eighty five percent purchase soap at least once every two months. Soap has somewhat low brand 172 loyalty compared to the other test product categories; it is a highly competitive category. The average woman in the sample buys over 25% of her soap on deal, having bought 2.6 of her last ten category purchases with a coupon. Brand recall and advertising—related recall showed significant main effects along repetition for both advertising and sales promotion. The differences between advertising and sales promotion, however, were not significant. The brand "value" had significant covariate effects for both brand loyalty and dealing which washed away the significance for the repetition effects on this variable. A significant interaction effect on brand evaluation occurred but no significance resulted along either of the main effects. Advertising became increasingly more positive and sales promotion became more negative with respect to brand evaluations (See Table 4.26). The overall cognitive response valence indexes in Table 4.27 show a repetition main effect for sales promotion; repetition on sales promotion decreased the overall favorable response so that it was almost neutral. This downward inflection for sales promotion seems to be reflected in both the advertising—related response and the product-related comments. Sales promotion in this category receives less positive comment in general, yet both advertising and sales promotion generate positive feedback from the sample. 173 TABLE 4.26 SOAP EFFECTS Exposures l 2 3 Total Prob. Brand Recall Total 67.6 90.5 89.4 82.8 .000* Advertising 70.7 88.2 90.9 83.2 .03* Sales Promotion 63.6 92.5 87.8 82.5 .003t Adv. vs S.P. 1.00 Related Recall Total 62.2 86.5 89.4 79.8 .000* Advertising 63.4 88.2 86.4 79.0 .01* Sales Promotion 60.6 85.0 92.7 80.7 .002* Adv. vs S.P. .87 Value Covariates: Purchase Frequency .76 Brand Loyalty .005X Coupon Experience .03* Total 6.51 7.42 6.72 6.87 Advertising 6.30 7.70 7.05 6.97 .08 Sales Promotion 6.77 7.17 6.35 6.76 .38 Interaction .42 Evaluation Covariates: Purchase Frequency .85 Brand Loyalty .04* Coupon Experience .11 Total 2.56 2.46 2.36 2.45 Advertising 2.76 2.33 2.27 2.45 .17 Sales Promotion 2.32 2.57 2.46 2.46 .99 Interaction .03* Purchase Intention Covariates: Purchase Frequency .62 Brand Loyalty .002* Coupon Experience .02* Total 5.90 6.31 6.06 6.09 Advertising 5.59 6.50 6.66 6.25 .75 Sales Promotion 6.26 6.14 5.41 5.91 .32 Interaction .16 174 TABLE 4.27 SOAP COGNITIVE RESPONSE Exposures l 2 3 Total Prob. Overall Total .21 .34 .13 .22 Advertising .24 .57 .22 .32 .19 Sales Promotion .16 .14 .03 .ll .03* Interaction .40 Product Total .12 .24 .13 .16 Advertising .12 .41 .20 .24 .32 Sales Promotion .12 .10 .05 .09 .053 Interaction .31 Advertising Total .05 .07 .00 .04 Advertising .05 .00 .02 .03 .58 Sales Promotion .06 .13 -.02 .05 .65 Interaction .45 Table 4.28 reveals very little difference in the recall of coupons between deal prone and non-deal prone persons for this advertisement. This difference is not significant. TABLE 4.28 SOAP COUPON MENTION BY DEAL PRONENESS Non—Deal Deal Prone Prone Total Prob. Percentage Recall 18.9 21.3 20.2 .928 CHAPTER V SUMMARY AND CONCLUSIONS Summary As a way of summarizing the data which has been presented, the individual product category findings will be reviewed together for each criterion variable and related back to the hypotheses outlined in Chapter II. Brand and Advertising Recall The first hypothesis concerns the brand and advertising- related recall measures for advertising versus sales promotion. Hypothesis 1: Brand and advertising related recall (learning) will be significantly different for ads with coupons than for ads without coupons. Brand recall was not different for advertising than for sales promotion significantly in any product category. Therefore, the research hypothesis is rejected and the null hypothesis is not. The general trends of the data vary according to the category, making any generalization difficult. Since all brands were on the market at the time of the study, it may be that the brand names were familiar to everyone enough to be remembered within a one-hour time frame. 175 176 Only one product category showed significant results along the advertising recall measure—-deodorant. In this category, advertising performed better than sales promotion over all three repetitions, although the gap closed on the third exposure. In the short term, the coupon in the sales promotion condition may be distracting from recall of the advertising content. For this first hypothesis, however, the data generally does not support the idea that sales promotion either attracts or detracts from recall of brand or a standard industry related recall. Brand Evaluation The second hypothesis was as follows: Hypothesis 2 Brand evaluation will be significantly different for ads with coupons than for ads without coupons. This hypothesis is also not supported with this data and the null is not rejected. Only the hair remover category came close to significance, with advertising performing best after one exposure and sales promotion increasing evaluation after two and three exposures. Several of the more frequently purchased product categories showed sales promotion to have consistently higher impact on this measure than advertising, but the differences were not significant. 177 Value and Purchase Intentions Hypothesis 3: Brand value and purchase intention will be different for those subjects who were exposed to an ad with a coupon versus those who were exposed to an ad without the coupon. Two categories were significant on the value criterion measure. Hair remover and nail polish remover, two of the infrequently purchased products but better-known brand names, revealed significantly higher overall value ratings under the sales promotion condition. Hair remover showed advertising to elicit the highest value at one exposure but sales promotion to significantly increase this value at higher exposure levels. Interestingly, the trends for the hair remover brand evaluation showed the exact same pattern, implying that a coupon in this case increases value and brand evaluation. Sales promotion produced significantly higher value ratings at all three exposure levels for the nail polish remover brand. However, this finding is exactly opposite the trends found in the evaluations for this brand. In other words, sales promotion for nail polish remover increased brand value but decreased overall evaluations. The drink mix, an expensive brand according to many women, received better value and brand ratings across sales promotion as compared to advertising, but these were not significant. Purchase intentions were not affected by the experimental conditions at all. This measure seemed relatively insensitive to promotion, perhaps due to the relatively short time period involved. 178 The research hypothesis is partially supported; the data suggests that for some product categories or brands, perhaps for dominant, infrequently purchased brands, sales promotion works well to increase brand value, although it is not possible to say how this affects purchase intention. The evidence regarding purchase intentions is inconclusive and does not support the hypothesis. Cognitive Responses The research hypotheses regarding the overall, product- related, and advertising-related cognitive responses upon exposure to brand advertising messages are tentatively supported depending on product category and/or advertising manipulations and the null rejected. They are listed once again, below. Hypothesis 4: Overall cognitive response to an advertisement with a coupon will be different in valence from an advertisement without a coupon. Hypothesis 5: Ad-related cognitive response to an advertisement with a coupon will be different in valence from an advertisement without a coupon. Hypothesis 6: Product—related cognitive response to an advertisement with a coupon will be different in valence from an advertisement without a coupon. Hair remover, once again, and soap were significant along the overall cognitive response variable; hair remover and nail polish remover were significant for advertising response; the drink mix was significant for the product 179 response and hair remover was very close to significant. Each of these will briefly be discussed individually. Hair remover showed significance for two of the three cognitive response indicators, but in all three cases, the sales promotion produced a net valence that was more positive than that for advertising. These results indicate evidence that a coupon in an advertisement may distract from the negative counterarguments against the product and the negative responses generated by a disliked advertising execution. It is worth noting that the valence for all indicators over both conditions was negative, therefore, sales promotion although eliciting negative thoughts, produced fewer negative comments as compared to advertising for both the product and creative execution. Since the only difference in the ad was the coupon price—off, the coupon either distracted attention away from the negative aspects of both the ad and product, or added enough positive thoughts to balance and exceed the negatives in both these response categories. Sales promotion, in this category, also improved value ratings and brand evaluation. In the soap category, responses in both conditions were positive overall. However, sales promotion produced less positive response for the product and the overall response. It appears that the coupon acted as a distraction in this advertisement, too, distracting from the positive support arguing that the advertising generated. These findings support those of the cognitive response and distraction research. Cognitive responses reflect and 180 probably influence the reception of the message and a distraction focuses away from the negative as well as the positive. The coupon helped the nail polish remover brand by distracting from a visual thought to be offensive and inhibiting counterarguing against this product. On the other hand, the coupon hurt the soap brand because it distracted from an advertisement most persons thought to be eye-catching and colorful and the favorable product attributes of this moisturizer soap brand. The drink mix category showed results similar to those of the soap category. Significant differences on the product—related response showed sales promotion to be positive and advertising to be negative. This pattern is mirrored in the advertising evaluation, although it is not significant. Some general observations about the responses generated by this particular brand will explain these findings. Subjects tended to view this product as expensive and disliked the health-oriented product concept because the drink was not nutritious and contained a sweetener. The advertising slogan was considered by many to be trite. Given this situation, a possible explanation is that the coupon distracted from these negative responses. Coupon Recall The coupon recall hypothesis was: Hypothesis 7: Those who use coupons will recall the presence of a coupon in the advertising at a different rate (greater) than those who do not use coupons. 181 Deal proneness did seem to affect the ability to recall or at least the tendency to mention the presence of a coupon in an advertisement, supporting this hypothesis. Only two categories showed significance on this variable, hair remover and mascara, although another four categories showed the same pattern of deal prone persons mentioning the coupon more often than the non—deal prone persons. Therefore, the null is rejected on two of the eight product categories but is not rejected for the others. It is notable that the coupons in the mascara and hair remover ads both were relatively obscure as compared to the coupons in other advertisements. Although not formulated into a hypothesis and not reported earlier, chi square analyses were highly significant (less than .001 in most cases) across repetition and promotion condition for coupon mentions. This is not surprising since it is unlikely that someone would mention the presence of a coupon when a coupon was never shown, and it is probable that repeated exposure to an advertisement with a coupon would increase the likelihood of remembering the coupon. However, what this does mean is that over the whole sample, both deal prone and non—deal prone, people do notice coupons. What is important here is that deal prone people are more likely to notice coupons all the time, even when they are underplayed in the advertise- ment, while people who do not use coupons, are less likely to notice them in general but in instances where the coupon is less obvious, are significantly less likely to notice the coupon. 182 Advertising Versus Sales Promotion Versus Both Hypothesis 8: Exposure to a mixed condition of both advertisements with and without coupons will affect all criterion variables differently than exposure to either alone at equal exposure levels. This eighth research hypothesis is rejected, leaving the null hypothesis as the explanation. The experimental conditions where both advertising and sales promotion were shown among the repetitions did not produce significant differences on any of the criterion variables measured. Furthermore, no discernible pattern was found. In almost every category, three exposures elicited the same level of response regardless of the mixture. These results are tabled for each product category across all measures in the appendices. When testing different promotion allocations more exposures may be necessary to document any differences. Brand and Advertising Recall over Repetition The brand and advertising recall were combined into one hypothesis which follows. Hypothesis 9: Brand and advertising related recall will differ over repetition levels. With respect to the first half of this hypothesis, the brand recall, it is well supported since seven of the eight product categories demonstrated significant repetition effects. Almost always recall increased positively across all three exposure levels, always with diminishing returns between the second and third exposures as compared to the In. PM 183 return from the first to the second exposure. In one category, nail polish remover (a factual black and white ad), the third repetition began to show an overall wearout or decline in recall. In general, then, repetition does seem to increase opportunity to learn about the product and to store the brand name in memory. Differences over repetition with respect to advertising recall were also evidenced. Six of the eight product categories showed significant increases over repetition in advertising recall. Again, diminishing returns were evident but wearout was displayed only for sales promotion in the nail polish remover category. Brand Evaluation, Brand Value, Purchase Intent, and Cognitive Response over Repetition None of these criterion measures showed any significant change over three repetitions of either sales promotion or advertising. Therefore, the last three hypotheses, stated on the next page, are not supported by this data and the corresponding null hypotheses are not rejected. Hypothesis 10: Brand evaluation will differ over repetition levels. Hypothesis 11: Brand value and purchase intentions will vary at different exposure levels. Hypothesis 12: Cognitive response will vary at different exposure levels. 184 The higher—level criterion measures do not seem to be sensitive to change in the short run after only three exposures in a laboratory. Although cognitive response does vary across exposure level, the incremental changes in net valence values are small and relatively gradual over the repetitions and, therefore, not significant. Conclusions The conclusions from this exploratory study are somewhat tentative. However, this study has proven useful for several reasons. It has shown the viability of laboratory research for studying sales promotion effects even in a very limited time period over a very few number of exposures. The results have suggested that learning theory and information processing theory may potentially offer valuable insight into differences in response between advertising and sales promotion, as well as among variations in creative approaches and design for both advertising and sales promotion. The study has also found the cognitive thought monitoring approach to be a very valuable measurement tool in research of this nature, perhaps of more utility than the usual outcome measures associated with the hierarchy of effects model. It has also shown the importance of product category, brand and execution differences in sales promotion research. Perhaps more importantly, this study represents another step toward understanding the role of one type of sales promotion--couponing-—in consumer learning processes and purchase decisions. Hopefully, future research will benefit 185 from both the positive findings of this research, but also from the problems encountered along the way. Future Research Directions In the Laboratory The potential of laboratory research is enormous in establishing cause-and—effect relationships between variables of interest or in investigating relatively complex events, such as consumer purchase decision making, at the developmental stage in a controlled environment. The primary advantage of laboratory experimentation is the ability to manipulate many variables simultaneously with absolute precision. Especially for advertising research, the ability to control exposure to promotional messages is important. This study has shown that product categories are very different from one another in many respects. Of concern to sales promotion research are the brand loyalty, purchase frequency and dealing characteristics of the category. Even among health and beauty aid products, the range of variation in these variables is wide. As an exploratory study, looking at several diverse categories was acceptable and helpful. However, in the future, it may be beneficial to select the categories in a more purposeful way to control for factors such as brand dominance, loyalty, purchase cycle and dealing. For instance, comparing several frequently purchased products with similar dealing patterns may reveal more consistency in response patterns. Comparing response across product types through careful manipulation and partitioning of variance, 186 may provide an understanding of how to better tailor the promotion program to the category. Experimenting across more product categories is also warranted. Totten (1985), in a paper on the health and beauty aid market, stated that these HBA products tend to be less price sensitive and promotion sensitive than other items. Grocery products, household products, drug products and pet products represent classes with heavy dealing activity where sales promotion response information would be welcome. Controlling for execution differences in the creative is also an important research consideration. Some of the differences in coupon recall and cognitive response to both product and advertising would seem to be able to be traced back to differences in the creative execution itself in this study. Information overload, which wasn’t sufficiently controlled in this experiment, is a theoretical area relevant to sales promotion. By controlling and manipulating the level of complexity, the mode of message presentation, the amount of clutter, and the cogency of the message and its relation to the price variable, some conclusions as to the role of a coupon in information overload may be discernible. Distraction theory holds promise for sales promotion effects research. However, the results of this study are not conclusive because either the product category or the creative and copy executions could account for the differences found between the sales promotion and advertising 187 conditions. Testing different executions for the same product would be necessary to address this issue. In this study, too, all coupons were treated as though they had the same attention-getting power. Obviously, such differences as the size of the coupon, the color of the coupon, the placement of the coupon, the face value, and other differences may influence the final response. Different coupon executions would be necessary to answer this question. In the Field Although laboratory research has proven valuable and will continue to be important in sales promotion research, field experimentation under controlled conditions is a natural next step. Field experiments not only offer more realistic exposure conditions, but allow a larger number of exposures to be tested over a longer period of time. In this exploratory study, the higher-level attitudinal and motivational measures were unaffected by the three exposures in the one-hour time period. Also, the question of whether advertising and sales promotion produce a synergistic effect when used together, remains unanswered. More exposures and a longitudinal design would allow direct monitoring of attitude change due to promotional mix. In past repetition research in the field, exposure control has been a problem that was addressed by using either direct mail techniques or very select media vehicles such as trade magazines. Since coupons are frequently delivered via 188 direct mail, this may be a good way to study their impact. Furthermore, redemption of the coupons, and product sales represent a behavioral measure not available in the laboratory. This study has dealt with variables which affect attitude development and change before a purchase is made. However, as Chapter II revealed, there are many theories with which to explain sales promotion effects. Price perception theories and post-purchase theories such as cognitive dissonance and attribution theory could be tested in the field given very special controlled conditions. Since it is probable that many purchase decisions are made in the store, especially for low involvement products, monitoring buying behavior in a real purchase situation would be helpful. Scanner panels, which provide both excellent purchase date, and which might also allow for in-home exposure to television advertising or direct mail coupons, are potentially a very good way to research sales promotion in the real world but in a somewhat controlled environment. APPENDIX A CAMPAIGN ADVERTISMENTS assess 5: .2: taxes 5.: so ‘ . $7.. . sea: a 32m cogency 56..“ EC .4. ._ .‘ 55 Lemon wars—m 2 EL ‘ < fies. Bee. as so .H Eagefifis ‘\. ”3.8 o S? . . > 03° 25.5 RM swag :wfim 53 ow * \ s APPENDIX B TEST ADVERTISMENTS 192 HAIR REMOVER ADVERTISEMENTS 81...... . .ggafldflfi Hz: .26—Shawl UZ—UDngPZL ., Bahia—$588.. .Egflcfizg assess... , its: . g88§§§£f§m§ :2 giaflaflg LEI «Hazfigl gébfiggqigsificfl. ., ._ 8. .352... =2.....z... _z_es.r...z,.ez_s§.. €355? liaise? “ Was—flu} N3...” slag... €35..£E..nu...... x . m“: 81.82 Eugenics... Pgéo.§§£§§ 193 NAIL POLISH REMOVER ADVERTISEMENTS 9.0 no. no 2 55:5: meU 3500 .0 ma 2090 on: 99.50. 05.5 05.29.: 0302.5 50 .0 030000 35: 50> 0. 05.29.: 600 020 0:0 55.90:. E 809 £00 55:5. 50 ESQ 2mm. 6:55.00 95 30.0503 2.295% Ego-.525) 5569.220 196.. 32 92:02 39: :02: .6: 5x05 :va: .E w mmz .25 5.7.3 :5 «2430!... 1 ,.= sea. 7‘6 AJmmua amman 555a: Eons. 550.0 65590.5 ~o<>=< Ema—cu: =23.— Eu .8 .n— was». 252.... :39. 5.0 _, .3503. r .1. _ — .05 no 02.028 _ can no. no u. 5.69:0”. .650 3.300 5 .ma _ 50cc... _ aims. 200E .9: 9.3. Sam .0 .moo £099 0. _ 29:. $2 9.5:. 0.0.0.9.: ..cm_=oo. 0.9.: .00 _ so: .255 95 an 55:3 .0 noon... 8 2005 09. 92.50. p.020 0550:. _ 029:: 50 .0 030000 «:0: 50> O. OSEOE _ mono 020 0:0 55.29.: c_ 9.09 £09 .9069. _ 50 06.0 mum. 5:83:00 95 305.000.. non! c0490» so: .88.: 2qu 922? to... 9:. 90.6 :25... £5 55...! 5.6 5,: 22 so: 59 925:. 99.. 63E :9. .o .8. .= was... :3 mg :2: 194 MASCARA ADVERTISEMENTS 3 9.....55. .8. 2:832 . ESE Eu :3 _ 26: 9.4 .852 3:532 m3 .0. 98mm... EOE 92:55 9953. :39... .2 .29.. 12.. r .39: Eu :8 _ .so: uc< .35.: 3:5qu as .2 583:. EOE “go—ca; 933.». L303 .2 5:9... Eu r. 195 HEADACHE REMEDY ADVERTISEMENTS mnwm - < mpflmo< .. U . com 3.23.3 com .3.“ 196 POWDERED DRINK MIX ADVERTISEMENTS 30> _o 50 $5.62 o 9.0:. __ = Ea: .03.».0 .S 30.0 o 3:060 v 32 m.w.m£ Dc< «.05: 6.20: 306:0“. .0 £0. c. 8E8 ED: 6.3.0 23:ch ,©c__mo_..om.o o «.905 302 197 SANITARY NAPKIN ADVERTISEMENTS gmwohuw .CEU .0 >153 8c m vac :30.» 5260.5 oEEEo. Eomvvflw 262 .0 53 E96 E .302 .mwfi m8H® 5.80 80:60.53 .Eowuofi >62 65...... .rkrileIEfmttr. Kim, .1 .. m ._u__ :3. 2c. RHHETHH coaom I:ll:fln..flnu.h “Om E «.8. , PM; .50: vc< . _ .89 c2: 30> 5...: , \ Eovvflm >52 Emu 6m . 50>.EoLBdmdme -flo... Jw EEO mo rung 09¢ m vac =.=o> Busvofi «ESE». Eovoflm >52 “0 x03 .605 E .302 1.3. $03.46 198 DEODORANT ADVERTISEMENTS Sin-15:1... fl ; . , .. .IIIJ W 111.1“-Jl'}! .ma.................. . ...:mm .3225... ll m. - 1 ,, . mm £5.53 a 9.: :9» $8.: .58 a 9:— 39» 3982.. M . m H" ZWZZWYKW »Zz< 20 oON m> and then circling the number of times you redeemed coupons within that amount.) Ever 922 992% Product Y_ee N_o Numbe; _f Coupeng o_ut g 19 Purchases Mascara U D 0 l 2 3 4 5 6 7 8 9 10 Hair Remover D C] 0 l 2 3 4 5 6 7 8 9 10 Nail Polish Remover [:1 D 0 l 2 3 4 5 6 7 8 9 10 Body Soap C] [:1 0 l 2 3 4 5 6 7 8 9 10 Powdered Drink Mix D D 0 l 2 3 4 5 6 7 8 9 10 Deodorant C] C] 0 1 2 3 4 5 6 7 8 9 10 Headache remedies C] D 0 1 2 3 4 5 6 7 8 9 10 Styling Mousse U U 0 l 2 3 4 5 6 7 8 9 10 Sanitary napkins C] D 0 l 2 3 4 5 6 7 8 9 10 Now, Just a few questions about you. What is your. age? ___________ years What is your marital status? [:1 Single DMarried D- Divorced D Other Where is your current local place of residence? [3 Dormitory [3 Apartment [3 House or condominium D Sorority or co-op D Other What is your current living situation? D Living alone [:1 Living with parents [3 Living with spouse [3 Living with roommate(s) D Other l4 216 Are you the principle purchaser of: Grocery products? D Yes D No Personal Items? [:1 Yes D No Are you a student at Michigan State University? C] Yes D No If yes, What is your major? If yes, What is your student status? D Graduate student [3 Senior [:1 Junior [:1 Sophomore E] Freshman On a typical day, about how many hours do you spend watching television? ______ ‘ On a typical day, about how many minutes do you spend reading magazines? ______ ~ # - For each of the following magazines, indicate whether you have read or glanced through the June or July issue, and then "X" the box which best describes the regularity with which you read this magazine. June or July issue? Isa 39 Llasxs 9:239 §21é9s 15222: Glamour C] D l 2 3 4 Cosmopolitan U D l 2 3 4 Vogue C] D 1 2 3 4 Self D D 1 2 3 4 Mademoiselle C] U 1 2 3 4 Harper’s Bazaar Cl C] l 2 3 4 Redbook C] C] l 2 3 4 Lady’s Home Journal U U l 2 3 4 McCall’s D E] 1 2 3 4 Good Housekeeping C] C] l 2 3 4 THANK YOU VERY MUCH FOR YOUR PARTICIPATION. 15 i J APPENDIX D COGNITIVE RESPONSE CATEGORY DESCRIPTIONS 217 PRODUCT/MESSAGE EVALUATIONS Counterarggments -- statements which are directed against the idea of or the use of the product in the advertising message and which: 1) state a specific unfavorable consequence of using the product, 2) state a specific undesirable attribute of the product, 3) suggest an alternative method for handling one of the problems cited in the advertising message, 4) state a specific favorable or desirable consequence or attribute of an alternative product, or which 5) challenge the accuracy or validity of a specific argument contained in the advertising message. These statements may take the form of declarative sentences or rhetorical questions. If the statement is in the form of a rhetorical question, its intent should be argumentative or express doubt or disbelief. The following types of statements are not to be considered as counterarguments: - simple statements of dislike for the product - emotional reactions which are not accompanied by any type of statement discussed above - statements falling into any of the other categories Su ort Ar uments —- statements which are directed in favor of the idea or use of the product in the adverti— sing message and which: 1) state a specific favorable consequence of using the product or a favorable reason for using the product, 2) state a specific desirable attribute of the product, 3) suggest an undesirable consequence of not using the product, or which 4) reaffirm the accuracy or validity of an argument presented in the advertisement. The following types of statements are not to be considered as support arguments: - simple statements of liking for the product - emotional reactions which are not accompanied by any type of statement discussed above - statements falling into any of the other categories 218 Simple Dieeffirgetiope -- statements which express: 1) a simple, unqualified statement expressing dislike or negative feelings about the product without offering any supporting reasons or explanations and without reference to particular product attributes, or 2) negative emotional reactions which are not accompanied by any other type of statements. Simple Affirmations -- statements which express: 1) a simple, unqualified statement expressing liking or positive feelings about the product without offering any supporting reasons or explanations and without reference to particular product attributes, or 2) positive emotional reactions which are not accompanied by any other type of statements. ADVERTISING EVALUATIONS Negative Advertieing Evaluations -- statements which are directed at the advertising approach or execution rather than the message or the product which express: 1) Posi 88 1) an unfavorable reaction to the advertisement. This includes statements such as, "The ad was offensive," or "I thought the model was ugly." dislike or disapproval for the approach taken by the advertiser in presenting the message such as,"I dislike comparative advertising (source derogation). tive Advertieing Evaluations -- statements which directed at the advertising approach or execution r rather than the message or the product which express: a favorable reaction to the advertisement. This includes statements such as, ”The ad was well done," or "I liked the coloring because it seemed feminine.” liking or approval for the approach taken by the advertiser in presenting the message (source bolstering). 10. 219 OTHER COGNITIVE RESPONSES Neutral Stategente -- statements toward either the advertising or the message/product which are neutral in valence. Simple repetitions of the message or observa- tions of something that was stated in the message or shown in the ad are considered to be neutral. Irrelevant Thogghte -- statements that do not reflect any type of relevant evaluation of the advertising message or the advertisement itself. These include: 1) any statements about one’s personal feeling or states such as "I wish this were over," or "I am getting tired of writing" 2) general statements about the task or the situation 3) statements which involve some extraneous thought generated by the task situation or the situations within the advertisements, such as "I remembered I need to buy some" Repetition—Related Evaluations -- statements which note that the advertising message was seen more than one time either within the experimental session itself or prior to the session such as, "The ad was shown several times," or "I’ve seen the television commercial for this brand many times." Non—User of Category -- statements which directly state that the individual does get use the product category (not the brand). APPENDIX E ANALYSIS OF VARIANCE SUMMARY TABLES FOR ADVERTISING AND SALES PROMOTION TABLE E.l r”) L. Recall Across all Product Categories Frequencies and Chi-Square Summaries for Brand HAIR REMOVER BRAND RECALL Advertising Sales Promotion 1 2 3 Total 1 2 3 Total No 11 7 15 33 13 14 15 42 Recall 28.2 21.2 37.5 29.5 33.3 35.9 36.6 35.3 Recall 28 26 25 79 26 25 26 77 71.8 78.8 62.5 70.5 66.7 64.1 63.4 64.7 Column 39 33 40 112 39 39 41 119 Total 34.8 29.5 35.7 100.0 32.8 32.8 34.5 100.0 Chi-Square D.F Sig. Repetition for Advertising 2.35386 2 3082 Repetition for Sales Promotion .10181 2 9504 Overall Repetition 1.23176 2 .5402 Overall Promotion .64822 1 4208 NAIL POLISH REMOVER BRAND RECALL Advertising Sales Promotion 1 2 3 Total 1 2 3 Total No 12 4 5 21 7 2 4 13 Recall 27.3 10.3 12.2 16.9 17.5 5.9 10.3 11.5 Recall 32 35 36 103 33 32 35 100 72.7 89.7 87.8 83.1 82.5 94.1 89.7 88.5 Column 44 39 41 124 40 34 39 113 Total 35.5 31.5 33.1 100.0 35.4 30.1 34.5 100.0 Chi—Square D.F. Sig. Repetition for Advertising 5.23401 2 .0730 Repetition for Sales Promotion 2.52756 2 2826 Overall Repetition 7.53287 2 0231 Overall Promotion 1.01162 1 .3145 221 TABLE 8.1 (Continued) HASCARA BRAND RECALL Advertising 1 2 3 Total No 6 2 1 9 Recall 15.4 5.9 2.6 8.0 Recall 33 32 38 103 84.6 94.1 97.4 92.0 Column 39 34 39 112 Total 34.8 30.4 34.8 100.0 Repetition for Advertising Repetition for Sales Promotion Overall Repetition Overall Promotion HEADACHE REMEDY BRAND RECALL Advertising 1 2 3 Total No 12 5 3 20 Recall 29.3 12.8 9.1 17.7 Recall 29 34 30 93 70.7 87.2 90.9 82.3 Column 41 39 33 113 Total 36.3 34.5 29.2 100.0 Repetition for Advertising Repetition for Sales Promotion Overall Repetition Overall Promotion Sales Promotion 1 2 3 Total 6 4 2 12 15.4 9.1 6.1 10.3 33 40 31 104 84.6 90.9 93.9 89.7 39 44 33 116 33.6 37.9 28.4 100.0 Chi—Square D.F. Sig. 4.64345 2 .0981 1.79571 2 .4074 5.96114 2 .0508 .13967 1 .7086 Sales Promotion 1 2 3 Total 14 8 6 28 42.4 20.5 14.0 24.3 19 31 37 87 57.6 79.5 86.0 75.7 33 39 43 115 28.7 33.9 37.4 100.0 Chi—Square D.F. Sig. 6.08330 2 .0478 8.68766 2 .0130 13.61167 2 .0011 1.14227 1 .2852 TABLE E.1 222 (Continued) POWDERED DRINK MIX BRAND RECALL No Recall Recall Column Total Advertising 1 2 3 Total 16 6 7 29 41.0 15.4 21.2 26.1 23 33 26 82 59.0 84.6 78.8 73.9 39 39 33 111 35.1 35.1 29.7 100.0 Repetition for Advertising Repetition for Sales Promotion Overall Repetition Overall Promotion Sales Promotion 1 2 3 19 7 5 44.2 17.1 12.5 24 34 35 55.8 82.9 87.5 43 41 40 34.7 33.1 32.3 Chi-Square D.F. 7.23018 2 13.14918 2 19.48642 2 .00229 1 Total 31 25.0 93 75.0 124 100.0 SANITARY NAPKIN No Recall Recall Column Total BRAND RECALL Advertising 1 2 3 Total 25 21 15 61 61.0 51.2 45.5 53.0 16 20 18 54 39.0 48.8 54.5 47.0 41 41 33 115 35.7 35.7 28.7 100.0 Repetition for Advertising Repetition for Sales Promotion Overall Repetition Overall Promotion Sales Promotion 1 2 3 26 23 15 66.7 53.5 38.5 13 20 24 33.3 46.5 61.5 39 43 39 32.2 35.5 32.2 Chi—Square D.F 1.85351 2 6.23547 2 7.43614 2 .00000 1 Total 64 52.9 57 47.1 121 100.0 223 TABLE E.1 (Continued) DEODORANT BRAND RECALL Advertising 1 2 3 Total No 19 17 13 49 Recall 57.6 39.5 31.7 41.9 Recall 14 26 28 68 42.4 60.5 68.3 58.1 Column 33 43 41 117 Total 28.2 36.8 35.0 100.0 Repetition for Advertising Repetition for Sales Promotion Overall Repetition Overall Promotion Sales Promotion 1 2 3 22 12 7 53.7 30.8 21.2 19 27 26 46.3 69.2 78.8 41 39 33 36.3 34.5 29.2 Chi-Square D.F. 5.18024 2 9.10928 2 13.26836 2 .53933 1 Total 41 36.3 72 63.7 113 100.0 SOAP BRAND RECALL Advertising 1 2 3 Total No 12 4 4 20 Recall 29.3 11.8 9.1 16.8 Recall 29 30 40 99 70.7 88.2 90.9 83.2 Column 41 34 44 119 Total 34.5 28.6 37.0 100.0 Repetition for Advertising Repetition for Sales Promotion Overall Repetition Overall Promotion Sales Promotion 1 2 3 12 3 5 36.4 7.5 12.2 21 37 36 63.6 92.5 87.8 33 40 41 28.9 35.1 36.0 Chi-Square D.F. 7.04528 2 11.67995 2 17.80521 2 .00000 1 Total 20 17.5 94 82.5 114 100.0 u \j- ‘- P. .- 224 TABLE 8.2 Frequencies and Chi-Square Summaries for Advertising-Related Recall Across all Product Categories HAIR REMOVER ADVERTISING RECALL Advertising Sales Promotion 1 2 3 Total 1 2 3 Total No 4 3 7 14 10 3 8 21 Recall 10.3 9.1 17.5 12.5 25.6 7.7 19.5 17.6 Recall 35 30 33 98 29 36 33 98 89.7 90.9 82.5 87.5 74.4 92.3 80.5 82.4 Column 39 33 40 112 39 39 41 119 Total 34.8 29.5 35.7 100.0 32.8 32.8 34.5 100.0 Chi—Square D.F. Sig. Repetition for Advertising 1.44442 2 .4857 Repetition for Sales Promotion 4.47238 2 .1069 Overall Repetition 3.79258 2 .1501 Overall Promotion .82231 1 .3645 NAIL POLISH REMOVER ADVERTISING RECALL Advertising Sales Promotion 1 2 3 Total 1 2 3 Total No 25 10 9 44 21 4 7 32 Recall 56.8 25.6 22.0 35.5 52.5 11.8 17.9 28.3 Recall 19 29 32 80 19 30 32 81 43.2 74.4 78.0 64.5 47.5 88.2 82.1 71.7 Column 44 39 41 124 40 34 39 113 Total 35.5 31.5 33.1 100.0 35.4 30.1 34.5 100.0 Chi-Square D.F. Sig. Repetition for Advertising 13.67833 2 .0011 Repetition for Sales Promotion 18.17834 2 .0001 Overall Repetition 30.77497 2 .0000 Overall Promotion 1.08389 1 .2978 225 TABLE E.2 (Continued) MASCARA ADVERTISING RECALL Advertising Sales Promotion l 2 3 Total 1 2 3 Total No 11 1 1 l3 9 5 4 18 Recall 28.2 2.9 2.6 11.6 23.1 11.4 12.1 15.5 Recall 28 33 38 99 30 39 9 98 71.8 97.1 97.4 88.4 76.9 88.6 87.9 84.5 Column 39 34 39 112 39 44 33 116 Total 34.8 30.4 34.8 100.0 33.6 37.9 28.4 100.0 Chi-Square D.F. Sig. Repetition for Advertising 16.06927 2 .0003 Repetition for Sales Promotion 2.56953 2 .2767 Overall Repetition 14.65846 2 0007 Overall Promotion .44609 1 .5042 HEADACHE REMEDY ADVERTISING RECALL Advertising Sales Promotion 1 2 3 Total 1 2 3 Total No 15 5 l 21 15 10 5 30 Recall 36.6 12.8 3.0 18.6 45.5 25.6 11.6 26.1 Recall 26 34 32 92 18 29 38 85 63.4 87.2 97.0 81.4 54.5 74.4 88.4 73.9 Column 41 39 33 113 33 39 43 115 Total 36.3 34.5 29.2 100.0 28.7 33.9 37.4 100.0 Chi-Square D.F. Sig Repetition for Advertising 14.91355 2 0006 Repetition for Sales Promotion 11.08614 2 .0039 Overall Repetition 23.68314 2 0000 Overall Promotion 1.44086 1 2300 226 TABLE E.2 (Continued) POWDERED DRINK MIX ADVERTISING RECALL Advertising 1 2 3 Total No 9 8 6 23 Recall 23.1 20.5 18.2 20.7 Recall 30 31 27 88 76.9 79.5 81.8 79.3 Column 39 39 33 111 Total 35.1 35.1 29.7 100.0 Chi Repetition for Advertising Repetition for Sales Promotion 3 Overall Repetition 3 Overall Promotion Sales Promotion 1 2 3 Total 11 6 4 21 25.6 14.6 10.0 16.9 32 35 36 103 74.4 85.4 90.0 83.1 43 41 40 124 34.7 33.1 32.3 100.0 ~Square D.F. Sig. .26232 2 .8771 .80704 2 .1490 .02025 2 .2209 .33077 1 .5652 SANITARY NAPKIN ADVERTISING RECALL Advertising 1 2 3 Total No 14 12 5 31 Recall 34.1 29.3 15.2 27.0 Recall 27 29 28 84 65.9 70.7 84.8 73.0 Column 41 41 33 115 Total 35.7 35.7 28.7 100.0 Chi Repetition for Advertising 3 Repetition for Sales Promotion 7 Overall Repetition 10 Overall Promoton Sales Promotion 1 2 3 Total 13 9 3 25 33.3 20.9 7.7 20.7 26 34 36 , 96 66.7 79.1 92.3 79.3 39 43 39 121 32.2 35.5 32.2 100.0 —Square D.F. Sig .52330 2 .1718 .82399 2 0200 .84775 2 .0044 .96673 1 3255 227 TABLE E.2 (Continued) DEODORANT ADVERTISING RECALL Advertising Sales Promotion 1 2 3 Total 1 2 3 Total No 17 11 7 35 27 18 6 51 Recall 51.5 25.6 17.1 29.9 65.9 46.2 18.2 45.1 Recall 16 32 34 82 14 21 27 62 48.5 74.4 82.9 70.1 34.1 53.8 81.8 54.9 Column 33 43 41 117 41 39 33 113 Total 28.2 36.8 35.0 100.0 36.3 34.5 29.2 100.0 Chi—Square D.F. Sig Repetition for Advertising 10.95390 2 0042 Repetition for Sales Promotion 16.80480 2 .0002 Overall Repetition 27.96012 2 0000 Overall Promotion 5.05519 1 0246 SOAP ADVERTISING RECALL Advertising Sales Promotion 1 2 3 Total 1 2 3 Total No 15 4 6 25 13 6 3 22 Recall 36.6 11.8 13.6 21.0 39.4 15.0 7.3 19.3 Recall 26 30 38 94 20 34 38 92 3.4 88.2 86.4 79.0 60.6 85.0 92.7 80.7 Column 41 34 44 119 33 40 41 114 Total 34.5 28.6 37.0 100.0 28.9 35.1 36.0 100.0 Chi—Square D.F. Sig. Repetition for Advertising 9.18642 2 0101 Repetition for Sales Promotion 12.81050 2 0017 Overall Repetition 21.22749 2 0000 Overall Promotion .02621 1 8714 228 TABLE E.3 Frequencies and Chi—Square Summaries for Coupon Recall According to Deal Proneness Categories HAIR REMOVER COUPON RECALL Not Prone Prone Total Do not 97 4 101 Mention 87.4 50.0 84.9 Mention 14 4 18 Coupon 12.6 50.0 15.1 Column 111 8 119 Total 93.3 6.7 100.0 Chi-Square = 5.47360 D.F. = 1 Significance = .0193 NAIL POLISH REMOVER COUPON RECALL Not Prone Prone Total Do not 40 12 52 Mention 51.9 33.3 46.0 Mention 37 24 61 Coupon 48.1 66.7 54.0 Column 77 36 113 Total 68.1 31.9 100.0 Chi-Square = 2.71346 D.F. = 1 Significance = .0995 MASCARA COUPON RECALL Not Prone Prone Total Do not 78 14 92 Mention 83.9 60.9 79.3 Mention l5 9 24 Coupon 16.1 39.1 20.7 Column 93 23 116 Total 80.2 19.8 100.0 Chi-Square = 4.62623 D.F. = 1 Significance = .0315 229 TABLE 8.3 (Continued) HEADACHE REMEDY COUPON RECALL Not Prone Prone Total Do not 63 35 98 Mention 87.5 81.4 85.2 Mention 9 8 l7 Coupon 12.5 18.6 14.8 Column 72 43 115 Total 62.6 37.4 100.0 Chi-Square = .38554 D.F Significance POWDERED DRINK MIX COUPON RECALL Not Prone Prone Total Do not 57 32 89 Mention 71.3 72.7 71.8 Mention 23 12 35 Coupon 28.8 27.3 28.2 Column 80 44 124 Total 64.5 35.5 100.0 Chi—Square = .00000 D.F. 1 Significance 1.0000 SANITARY NAPKIN COUPON RECALL Not Prone Prone Total Do not 54 43 97 Mention 77.1 84.3 80.2 Mention 16 8 24 Coupon 22.9 15.7 19.8 Column 70 51 121 Total 57.9 42.1 100.0 Chi-Square = .55645 D.F Significance ill 230 TABLE 8.3 (Continued) DEODORANT COUPON RECALL Not Prone Prone Total Do not 51 48 99 Mention 92.7 82.8 87.6 Mention 4 10 14 Coupon 7.3 17.2 12.4 Column 55 58 113 Total 48.7 51.3 100.0 Chi—Square = 1.74770 D.F. = Significance = .1862 SOAP COUPON RECALL Not Prone Prone Total Do not 43 48 91 Mention 81.1 78.7 79.8 Mention 10 13 23 Coupon 18.9 21.3 20.2 Column 53 61 114 Total 46.5 53.5 100.0 Chi-Square = .00815 D.F. = Significance = .9280 231 TABLE 8.4 Analysis of Variance Summary Tables for Brand Value Across Product Categories HAIR REMOVER BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 8.605 3 2.868 .462 .710 Purchase Frequency 3.432 1 3.432 .553 .460 Brand Loyalty 1.785 1 1.785 .288 .594 Coupon Usage 3.814 1 3.814 .614 .437 Main Effects 30.530 3 10.177 1.639 .191 Promotion 29.668 1 29.668 4.779 .033 Repetition 2.063 2 1.031 .166 .847 2-way Interactions 25.344 2 12.672 2.041 .140 Explained 64.479 8 8.060 1.298 .264 Residual 335.235 54 6.208 Total 399.714 62 6.447 NAIL POLISH REMOVER BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 82.825 3 27.608 7.228 .000 Purchase Frequency 31.018 1 31.018 8.121 .005 Brand Loyalty 17.882 1 17.882 4.682 .032 Coupon Usage 8.072 1 8.072 2.113 .147 Main Effects 35.770 3 11.923 3.122 .027 Promotion 18.324 1 18.324 4.798 .030 Repetition 18.030 2 9.015 2.360 .097 2-way Interactions 1.599 2 .799 .209 .811 Explained 120.194 8 15.024 3.934 .000 Residual 859.379 225 3.819 Total 979.573 233 4.204 MASCARA BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 42.694 3 14.231 2.304 .078 Purchase Frequency 5.536 1 5.536 .896 .345 Brand Loyalty 12.040 1 12.040 1.949 .164 Coupon Usage 21.813 1 21.813 3.531 .062 Main Effects 1.591 3 .530 .086 .968 Promotion .304 1 .304 .049 .825 Repetition 1.211 2 .605 .098 .907 2-way Interaction .329 2 .164 .027 .974 Explained 44.614 8 5.577 .903 .515 Residual 1216.944 197 6.177 Total 1261.558 N O 01 6.154 TABLE 8.4 (Continued) 232 HEADACHE REMEDY BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 3.402 3 1.134 .253 .859 Purchase Frequency 1.294 1 1.294 .289 .592 Brand Loyalty 1.476 1 1.476 .330 .567 Coupon Usage .605 1 .605 .135 .714 Main Effects 21.863 3 7.288 1.627 .186 Promotion 13.303 1 13.303 2.970 .087 Repetition 8.542 2 4.271 .953 .388 2—way Interactions 1.915 2 .958 .214 .808 Explained 27.180 8 3.398 .758 .640 Residual 645.029 144 4.479 Total 672.209 152 4.422 POWDERED DRINK MIX BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 109.981 3 36.660 4.866 .003 Purchase Frequency 49.080 1 49.080 6.515 .012 Brand Loyalty 39.329 1 39.329 5.220 .024 Coupon Usage 4.196 1 4.196 557 .457 Main Effects 12.534 3 4.178 555 .646 Promotion 6.031 1 6.031 800 .373 Repetition 6.270 2 3.135 416 .661 2—way Interactions 2.296 2 1.148 .152 .859 Explained 124.811 8 15.601 2.071 .045 Residual 783.508 104 7.534 Total 908.319 112 8.110 SANITARY NAPKIN BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 49.202 3 16.401 4.428 .005 Purchase Frequency 7.714 1 7.714 2.083 .151 Brand Loyalty 4.168 1 4.168 1.125 .290 Coupon Usage 36.552 1 36.552 9.869 .002 Main Effects 6.783 3 2.261 .610 .609 Promotion .561 1 .561 .151 .698 Repetition 5.897 2 2.949 .796 .453 2-way Interactions 12.720 2 6.360 1.717 .183 Explained 68.705 8 8.588 2.319 .022 Residual 622.244 168 3.704 Total 690.949 176 3.926 233 TABLE 8.4 (Continued) DEODORANT BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 20.031 3 6.677 1.388 .248 Purchase Frequency 1.135 1 1.135 .236 .628 Brand Loyalty 19.162 1 19.162 3.984 .047 Coupon Usage .001 1 .001 .000 .986 Main Effects 6.935 3 2.312 .481 .696 Promotion 1.344 1 1.344 .279 .598 Repetition 5.800 2 2.900 .603 .548 2-way Interactions 10.922 2 5.461 1.136 .324 Explained 37.888 8 4.736 .985 .449 Residual 856.027 178 4.809 Total 893.914 186 4.806 SOAP BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 62.209 3 20.736 4.406 .005 Purchase Frequency .461 1 .461 .098 .755 Brand Loyalty 37.798 1 37.798 8.030 .005 Coupon Usage 22.424 1 22.424 4.764 .030 Main Effects 26.992 3 8.997 1.912 .129 Promotion 3.692 1 3.692 .784 .377 Repetition 24.419 2 12.209 2.594 .077 2-way Interactions 8.306 2 4.153 .882 .415 Explained 97.507 8 12.188 2.590 .010 Residual 950.777 202 4.707 Total 1048.284 210 4.992 —-——-—----——_----‘-———-----u—--‘_——_—-——**—_———---—--—————-—— 234 TABLE 8.5 Analysis of Variance Summary Tables for Brand Evaluation Across Product Categories HAIR REMOVER BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 4.436 3 1.479 1.984 .127 Purchase Frequency 3.463 1 3.463 4.647 .036 Brand Loyalty .015 l .015 .020 .889 Coupon Usage .269 1 .269 .360 .551 Main Effects 4.464 3 1.488 1.997 .125 Promotion 2.743 1 2.743 3.681 .060 Repetition 1.746 2 .873 1.171 .318 2-way Interactions 4.130 2 2.065 2.771 .071 Explained 13.030 8 1.629 2.186 .043 Residual 40.240 54 .745 Total 53.270 62 .859 NAIL POLISH REMOVER BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 2.350 3 .783 1.827 .143 Purchase Frequency .149 1 .149 .347 .556 Brand Loyalty 1.501 1 1.501 3.502 .063 Coupon Usage .123 l .123 .286 .593 Main Effects 2.564 3 .855 1.994 .116 Promotion .702 1 .702 1.637 .202 Repetition 1.844 2 .922 2.151 .119 2-way Interactions .032 2 .016 .037 .963 Explained 4.946 8 .618 1.442 .180 Residual 96.460 225 .429 Total 101.406 233 .435 MASCARA BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 10.542 3 3.514 4.837 .003 Purchase Frequency 6.951 1 6.951 9.569 .002 Brand Loyalty .035 l .035 .048 .826 Coupon Usage 3.538 1 3.538 4.871 .028 Main Effects .629 3 .210 .288 .834 Promotion .025 1 .025 .035 .852 Repetition .606 2 .303 .417 .659 2-way Interactions .550 2 .275 .379 .685 Explained 11.721 8 1.465 2.017 .046 Residual 147.468 203 .726 Total 159.189 211 .754 235 TABLE 8.5 (Continued) HEADACHE REMEDY BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 1.992 3 .664 1.189 .316 Purchase Frequency .451 1 .451 .808 .370 Brand Loyalty .348 1 .348 .624 .431 Coupon Usage .959 1 .959 1.717 .192 Main Effects .593 3 .198 .354 .786 Promotion .512 1 .512 .918 .340 Repetition .144 2 .072 .129 .879 2-way Interactions .065 2 .032 .058 .944 Explained 2.650 8 .331 .593 .782 Residual 80.409 144 .558 Total 83.059 152 .546 POWDERED DRINK MIX BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 13.073 3 4.358 6.408 .001 Purchase Frequency 3.558 1 3.558 5.231 .024 Brand Loyalty 6.457 1 6.457 9.494 .003 Coupon Usage .747 l .747 1.098 .297 Main Effects 2.515 3 .838 1.233 .302 Promotion .794 1 .794 1.167 .283 Repetition 1.789 2 .895 1.315 .273 2-way Interactions .621 2 .310 .456 .635 Explained 16.209 8 2.026 2.979 .005 Residual 70.729 104 .680 Total 86.938 112 .776 SANITARY NAPKINS BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 7.638 3 2.546 5.026 .002 Purchase Frequency .760 1 .760 1.500 .222 Brand Loyalty 1.081 1 1.081 2.134 .146 Coupon Usage 5.548 1 5.548 10.951 .001 Main Effects .425 3 .142 .280 .840 Promotion .012 1 .012 .024 .877 Repetition .414 2 .207 .408 .665 2-way Interactions .346 2 .173 .342 .711 Explained 8.410 8 1.051 2.075 .041 Residual 85.116 168 .507 Total 93.525 176 .531 23 TABLE 8.5 (Continued) 6 DEODORANT BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 1.704 3 .568 .930 .427 Purchase Frequency .205 1 .205 .335 .563 Brand Loyalty 1.412 1 1.412 2.314 .130 Coupon Usage .079 1 .079 .129 .719 Main Effects 1.213 3 .404 .662 .576 Promotion .538 1 .538 .881 .349 Repetition .655 2 .328 .537 .586 2-way Interactions .469 2 .235 .384 .681 Explained 3.386 8 .423 .693 .697 Residual 108.657 178 .610 Total 112.043 186 .602 SOAP BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 3.524 3 1.175 2.342 .074 Purchase Frequency .019 1 .019 .038 .846 Brand Loyalty 2.129 1 2.129 4.246 .041 Coupon Usage 1.287 1 1.287 2.566 .111 Main Effects 1.797 3 .599 1.194 .313 Promotion .000 1 .000 .000 .985 Repetition 1.796 2 .898 1.791 .169 2—way Interactions 3.700 2 1.850 3.689 .027 Explained 9.021 8 1.128 2.249 .025 Residual 101.301 202 .501 Total 110.322 210 .525 237 TABLE 8.6 Analysis of Variance Summary Tables for Brand Purchase Intent Across Product Categories HAIR REMOVER PURCHASE INTENT a DP mNNHwHHI-Jw 1.211 3.445 .067 .022 .957 2.867 .025 .942 1.049 Signif of F .314 .069 .797 .883 .420 .096 .975 .396 .412 Sum of Source of Variation Square Covariates 32.618 Purchase Frequency 30.922 Brand Loyalty .599 Coupon Usage .197 Main Effects 25.772 Promotion 25.727 Repetition .456 2-way Interactions 16.915 Explained 75.305 Residual 484.632 Total 559.937 NAIL POLISH REMOVER PURCHASE Sum of Source of Variation Square Covariates 385.790 Purchase Frequency 90.563 Brand Loyalty 134.447 Coupon Usage 39.816 Main Effects 26.137 Promotion 8.886 Repetition 17.792 2-way Interactions 1.759 Explained 413.686 Residual 1118.199 Total 1531.885 INTENT D vs: mNNHwHo—H—Iw N N 01 233 128.597 90.563 134.447 39.816 8.712 8.886 8.896 .880 51.711 4.970 6.575 25.876 18.223 27.053 .012 .753 .788 .790 .177 10.405 v-u-II—‘m Signif of F .000 .000 .000 .005 .157 .183 .169 .838 .000 MASCARA PURCHASE INTENT Sum of Source of Variation Squares Covariates 114.582 Purchase Frequency 16.168 Brand Loyalty 44.305 Coupon Usage 44.193 Main Effects 7.592 Promotion 2.865 Repetition 4.298 2-way Interactions 6.752 Explained 128.926 Residual 1488.627 Total 1617.553 mNNI—th—owww .054 .140 .863 .848 .335 .379 .284 .447 2.133 OIOINO'I Signif of F .002 .145 .016 .017 .800 .539 .753 .640 .034 TABLE 8.6 (Continued) 238 HEADACHE REMEDY PURCHASE INTENT Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 9.405 3 3.135 .392 .759 Purchase Frequency 5.673 1 5.673 .709 .401 Brand Loyalty 2.740 1 2.740 .343 .559 Coupon Usage .370 l .370 .046 .830 Main Effects 19.981 3 6.660 .833 .478 Promotion 3.636 1 3.636 .455 .501 Repetition 18.184 2 9.092 1.137 .324 2-way Interactions 20.634 2 10.317 1.290 .278 Explained 50.019 8 6.252 .782 .619 Residual 1151.510 144 7.997 Total 1201.529 152 7.905 POWDERED DRINK MIX PURCHASE INTENT Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 220.831 3 73.610 9.321 .000 Purchase Frequency 54.437 1 54.437 6.893 .010 Brand Loyalty 123.959 1 123.959 15.696 .000 Coupon Usage 8.211 1 8.211 1.040 .310 Main Effects 23.027 3 7.676 .972 .409 Promotion 3.945 1 3.945 .500 .481 Repetition 19.657 2 9.828 1.245 .292 2-way Interactions 9.692 2 4.846 .614 .543 Explained 253.550 8 31.694 4.013 .000 Residual 821.317 104 7.897 Total 1074.867 112 9.597 SANITARY NAPKIN PURCHASE INTENT Sum of Mean Signif Source of Variation Squares DF Square F of F Covariates 184.397 3 61.466 7.298 .000 Purchase Frequency 13.816 1 13.816 1.640 .202 Brand Loyalty 12.741 1 12.741 1.513 .220 Coupon Usage 153.805 1 153.805 18.261 .000 Main Effects 11.650 3 3.883 .461 .710 Promotion 2.799 1 2.799 .332 .565 Repetition 8.048 2 4.024 .478 .621 2-way Interactions 23.493 2 11.746 1.395 .251 Explained 219.540 8 27.442 3.258 .002 Residual 1415.002 168 8.423 Total 1634.542 176 9.287 239 TABLE 8.6 (Continued) DEODORANT PURCHASE INTENT mNNHwHHt-aw H «q d) 186 3.074 .503 8.139 .368 .841 .112 1.192 .459 1.583 Signif of F .029 .479 .005 .545 .473 .739 .306 .632 .133 Sum of Source of Variation Squares Covariates 87.619 Purchase Frequency 4.779 Brand Loyalty 77.336 Coupon Usage 3.497 Main Effects 23.986 Promotion 1.060 Repetition 22.652 2-way Interactions 8.729 Explained 120.335 Residual 1691.441 Total 1811.775 SOAP PURCHASE INTENT Sum of Source of Variation Squares Covariates 111.955 Purchase Frequency 1.780 Brand Loyalty 69.313 Coupon Usage 38.361 Main Effects 10.606 Promotion 6.984 Repetition 4.070 2-way Interactions 26.427 Explained 148.988 Residual 1427.476 Total 1576.464 DF mNNHwHt-ah-Iw N O N 210 5.281 .252 9.808 5.428 .500 .988 .288 1.870 2.635 Signif of F .002 .616 .002 .021 .683 .321 .750 .157 .009 240 TABLE 8.7 Analysis of Variance Summary Tables for Overall Cognitive Response Across Product Categories HAIR REMOVER OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 4.068 3 1.356 3.401 .024 Promotion 3.709 1 3.709 9.302 .003 Repetition .298 2 .149 .374 .690 2—way Interactions .476 2 .238 .596 .554 Explained 4.544 5 .909 2.279 .059 Residual 22.726 57 .399 Total 27.270 62 .440 NAIL POLISH REMOVER OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 1.252 3 .417 .780 .506 Promotion .124 1 .124 .232 .630 Repetition 1.123 2 .561 1.049 .352 2-way Interactions 2.112 2 1.056 1.973 .141 Explained 3.363 5 .673 1.257 .284 Residual 122.021 228 .535 Total 125.385 233 .538 MASCARA OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 2.032 3 .677 .772 .511 Promotion 1.266 1 1.266 1.442 .231 Repetition .882 2 .441 .502 .606 2-way Interactions 2.305 2 1.153 1.313 .271 Explained 4.338 5 .868 .989 .426 Residual 175.507 200 .878 Total 179.845 205 .877 HEADACHE REMEDY OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .924 3 .308 .546 .651 Promotion .147 l .245 .406 .525 Repetition .762 2 .381 .676 .510 2-way Interactions 1.155 2 .577 1.024 .362 Explained 2.079 5 .416 .737 .597 Residual 83.460 148 .564 Total 85.539 153 .559 241 TABLE 8.7 (Continued) POWDERED DRINK MIX OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Squares F of F Main Effects 3.075 3 1.025 1.311 .271 Promotion 2.045 1 2.045 2.616 .107 Repetition 1.080 2 .540 .691 .502 2-way Interactions .221 2 .110 .141 .868 Explained 12.674 8 1.584 2.026 .044 Residual 175.112 224 .782 Total 187.785 232 .809 SANITARY NAPKIN OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 2.728 3 .909 1.102 .350 Promotion 1.328 1 1.328 1.610 .206 Repetition 1.637 2 .818 .992 .373 2-way Interactions .324 2 .162 .197 .822 Explained 3.053 5 .611 .740 .595 Residual 141.117 171 .825 Total 144.169 176 .819 DEODORANT OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 3.225 3 1 075 1.944 .124 Promotion .267 l 267 .482 .488 Repetition 2.997 2 1.498 2.710 .069 2-way Interactions .552 2 .276 .499 .608 Explained 3.777 5 .755 1.366 .239 Residual 100.084 181 .553 Total 103.861 186 .558 SOAP OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 4.314 3 1.438 2.664 .049 Promotion 2.732 1 2.732 5.061 .026 Repetition 1.825 2 .912 1.690 .187 2-way Interactions 1.004 2 .502 .930 .396 Explained 5.317 5 1.063 1.970 .084 Residual 110.654 205 .540 Total 115.972 210 .552 242 TABLE 8.8 Analysis of Variance Summary Tables for Product- Related Cognitive Response Across Categories HAIR REMOVER PRODUCT-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 1.437 3 .479 1.789 .150 Promotion, .889 1 .889 3.319 .070 Repetition .512 2 .256 .957 .386 2-way Interactions .595 2 .297 1.111 .331 Explained 2.032 5 .406 1.518 .185 Residual 60.237 225 .268 Total 62.268 230 .271 NAIL POLISH REMOVER PRODUCT-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .988 3 .329 .809 .490 Promotion .162 l .162 .398 .529 Repetition .817 2 .409 1.004 .368 2-way Interactions .479 2 .239 .588 .556 Explained 1.466 5 .293 .721 .608 Residual 93.985 231 .407 Total 95.451 236 .404 MASCARA PRODUCT-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .904 3 .301 .503 .680 Promotion .826 1 .826 1.380 .241 Repetition .124 2 .062 .104 .901 2-way Interactions 1.771 2 .885 1.478 .230 Explained 2.675 5 .535 .893 .486 Residual 132.992 222 .599 Total 135.667 227 .598 HEADACHE REMEDY PRODUCT-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .968 3 .323 .701 .552 Promotion .403 1 .403 .875 .351 Repetition .602 2 .301 .654 .521 2-way Interactions .886 2 .443 .962 .384 Explained 1.854 5 ~ .371 .806 .547 Residual 102.159 222 .460 Total 104.013 227 .458 TABLE E.8 (Continued) POWDERED DRINK MIX PRODUCT-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 3.349 3 1.116 2.415 .067 Promotion 1.964 1 1.964 4.249 .040 Repetition 1.386 2 .693 1.499 .226 2-way Interactions .355 2 .178 .384 .681 Explained 3.705 5 .741 1.603 .160 Residual 105.870 229 .462 Total 109.574 234 .468 SANITARY NAPKIN PRODUCT-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 3.335 3 1.112 2.136 .096 Promotion .057 1 .057 .110 .740 Repetition 3.248 2 1.624 3.121 .046 2-way Interactions .437 2 .218 .420 .658 Explained 3.771 5 .754 1.449 .208 Residual 119.699 230 .520 Total 123. DEODORANT PRODUCT-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 2.302 3 .767 2.582 .054 Promotion .034 1 .034 .113 .737 Repetition 2.278 2 1.139 3.831 .023 2-way Interactions 1.004 2 .502 1.689 .187 Explained 3.307 5 .661 2.225 .053 Residual 66.589 224 .297 Total 69.896 229 .305 SOAP PRODUCT-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 2.114 3 .705 1.886 .133 Promotion 1.415 1 1.415 3.787 .053 Repetition .846 2 .423 1.132 .324 2—way Interactions .887 2 .443 1.187 .307 Explained 3.000 5 .600 1.606 .159 Residual 84.802 227 .374 Total 87.803 232 .378 244 TABLE 8.9 Analysis of Variance Summary Tables for Advertising-Related Cognitive Response HAIR REMOVER ADVERTISING-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 2.014 3 .671 2.185 .091 Promotion 1.350 1 1.350 4.395 .037 Repetition .731 2 .365 1.189 .307 2-way Interactions .780 2 .390 1.269 .283 Explained 2.794 5 .559 1.818 .110 Residual 69.137 225 .307 Total 71.931 230 .313 NAIL POLISH REMOVER ADVERTISING-RELATED COGNITIVE RESPONSE Sum of - Mean Signif Source of Variation Squares DF Square F of F Main Effects .502 2 .251 2.091 .127 Promotion .007 1 .007 .062 .803 Repetition .493 l .493 4.109 .044 2-way Interactions .075 1 .075 .624 .431 Explained .577 3 .192 1.602 .191 Residual 18.366 153 .120 Total 18.943 156 .121 MASCARA ADVERTISING-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .531 3 .177 1.036 .377 Promotion .381 1 .381 2.232 .137 Repetition .191 2 .095 .559 .573 2-way Interactions .005 2 .002 .014 .986 Explained .536 5 .107 .627 .679 Residual 37.933 222 .171 Total 38.469 227 .169 HEADACHE REMEDY ADVERTISING-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .535 3 .178 1.559 .200 Promotion .008 l .008 .072 .788 Repetition .517 2 .259 2.261 .107 2-way Interactions .045 2 .022 .195 .823 Explained .580 5 .116 1.013 .411 Residual 25.403 222 .114 Total 25.982 227 .114 TABLE 8.9 (Continued) POWDERED DRINK MIX ADVERTISING-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .369 3 .123 .491 .689 Promotion .090 1 .090 .358 .550 Repetition .273 2 .137 .546 .580 2—way Interactions .190 2 .095 .380 .684 Explained .559 5 .112 .447 .815 Residual 57.288 229 .250 Total 57.847 234 .247 SANITARY NAPKIN ADVERTISING-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 1.370 3 .457 1.967 .120 Promotion .572 1 .572 2.462 .118 Repetition .840 2 .420 1.809 .166 2-way Interactions .033 2 .016 .071 .932 Explained 1.402 5 .280 1.208 .306 Residual 53.390 230 .232 Total 54.792 235 .233 DEODORANT ADVERTISING-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .034 3 .011 .065 .978 Promotion .005 1 .005 .031 .860 Repetition .029 2 .015 .084 .919 2-way Interactions .354 2 .177 1.027 .360 Explained .387 5 .077 .450 .813 Residual 38.573 224 .172 Total 38.961 229 .170 SOAP ADVERTISING-RELATED COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .246 3 .082 .442 .723 Promotion .040 1 .040 .213 .645 Repetition .202 2 .101 .545 .581 2-way Interactions .297 2 .149 .801 .450 Explained .543 5 .109 .586 .711 Residual 42.109 227 .186 Total 42.652 232 .184 APPENDIX F ANALYSIS OF VARIANCE SUMMARY TABLES FOR ADVERTISING AND SALES PROMOTION MIX 246 Frequencies and Chi-Square Summaries for Brand Recall by Product Category TABLE F.1. HAIR REMOVER--BRAND RECALL 3-0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 15 13 16 15 59 Recall 37.5 31.7 36.4 36.6 35.5 Recall 25 28 28 26 107 62.5 68.3 63.6 63.4 64.5 Column 40 41 44 41 166 Total 24.1 24.7 26.5 24.7 100.0 Chi-Square = .36255 D.F. = 3 Significance = 9479 NAIL POLISH REMOVER--BRAND RECALL 3-0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 5 3 2 4 14 Recall 12.2 7.3 4.7 10.3 8.5 Recall 36 38 41 35 150 87.8 92.7 95.3 89.7 91.5 Column 41 41 43 39 164 Total 25.0 25.0 26.2 23.8 100.0 Chi-Square = 1.76010 D.F. = 3 Significance = 6237 MASCARA--BRAND RECALL 3-0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No l 3 5 2 11 Recall 2.6 7.3 12.5 6.1 7.2 Recall 38 38 35 31 142 97.4 92.7 87.5 93.9 92.8 Column 39 41 40 33 153 Total 25.5 26.8 26.1 21.6 100.0 Chi-Square = 3.00504 D.F Significance 3908 247 TABLE F.1. (Continued) HEADACHE REMEDY--BRAND RECALL 3-0 Mix 2-1 Mix 1—2 Mix 0—3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 3 5 5 6 19 Recall 9.1 11.4 12.2 14.0 11.8 Recall 30 39 36 37 142 90.9 88.6 87.8 86.0 88.2 Column 33 44 41 43 161 Total 20.5 27.3 25.5 26.7 100.0 Chi-Square = .43847 D.F. = 3 Significance = .9322 POWDERED DRINK MIX--BRAND RECALL 3-0 Mix 2—1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 7 3 1 5 16 Recall 21.2 7.7 2.9 12.5 11.0 Recall 26 36 33 35 130 78.8 92.3 97.1 87.5 89.0 Column 33 39 34 40 146 Total 22.6 26.7 23.3 27.4 100.0 Chi-Square = 6.31901 D.F. = 3 Significance = .0971 SANITARY NAPKIN--BRAND RECALL 3~0 Mix 2—1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 15 17 24 15 71 Recall 45.5 51.5 61.5 38.5 49.3 Recall 18 16 15 24 73 54.5 48.5 38.5 61.5 50.7 Column 33 33 39 39 144 Total 22.9 22.9 27.1 27.1 100.0 Chi-Square = 4.42995 D.F. = 3 Significance = .2186 248 TABLE F.1. (Continued) DEODORANT--BRAND RECALL 3-0 Mix 2—1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 13 7 13 7 40 . Recall 31.7 17.9 33.3 21.2 26.3 Recall 28 32 26 26 112 68.3 82.1 66.7 78.8 73.7 Column 41 39 39 33 152 Total 27.0 25.7 25.7 21.7 100.0 Chi—Square = 3.45646 D.F. = 3 Significance — 3265 SOAP--BRAND RECALL 3—0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 4 2 3 5 14 Recall 9.1 4.7 7.7 12.2 8.4 Recall 40 41 36 36 153 90.9 95.3 92.3 87.8 91.6 Column 44 43 39 41 167 Total 26.3 25.7 23.4 24.6 100.0 Chi-Square 1.60840 D.F Significance 6575 249 TABLE F.2. Frequencies and Chi-Square Summaries for Advertising-Related Recall by Product Category HAIR REMOVER--ADVERTISINC-RELATED RECALL 3-0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 7 3 5 8 23 Recall 17.5 7.3 11.4 19.5 13.9 Recall 33 38 39 33 143 82.5 92.7 88.6 80.5 86.1 Column 40 41 44 41 166 Total 24.1 24.7 26.5 24.7 100.0 Chi-Square 3.24173 D.F. = Significance 3558 NAIL POLISH REMOVER--ADVERTISING-RELATED RECALL 3-0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 9 12 7 7 35 Recall 22.0 29.3 16.3 17.9 21.3 Recall 32 29 36 32 129 78.0 70.7 83.7 82.1 78.7 Column 41 41 43 39 164 Total 25.0 25.0 26.2 23.8 100.0 Chi-Square = 2.46763 D.F. = 3 Significance - 4812 MASCARA--ADVERTISINC-RELATED RECALL 3-0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 1 2 3 4 10 Recall 2.6 4.9 7.5 12.1 6.5 Recall 38 39 37 29 143 97.4 95.1 92.5 87.9 93.5 Column 39 41 40 33 153 Total 25.5 26.8 26.1 21.6 100.0 Chi-Square 2.93768 D.F = Significance = .4013 TABLE F.2. (Continued) HEADACHE REMEDY--ADVERTISINC-RELATED RECALL 3—0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 1 8 10 5 24 Recall 3.0 18.2 24.4 11.6 14.9 Recall 32 36 31 38 137 97.0 81.8 75.6 88.4 85.1 Column 33 44 41 43 161 Total 20.5 27.3 25.5 26.7 100.0 Chi—Square = 7.31297 D.F. = 3 Significance = 0626 POWDERED DRINK MIX--ADVERTISING-RELATED RECALL Column Total Chi-Square SANITARY NAPKIN--ADVERTISING-RELATED RECALL No Recall Recall Column Total Chi-Square 3—0 Mix (3 Ads) 6 18.2 27 81.8 33 22.6 2.38739 3-0 Mix (3 Ads) 5 15.2 28 84.8 33 22.9 2.18981 2-1 Mix (2 Ads) 3 7.7 36 92.3 39 26.7 2-1 Mix (2 Ads) 6 18.2 27 81.8 33 22.9 0-3 Mix' Total 1-2 Mix (2 SP) (3 SP) 3 4 16 8.8 10.0 11.0 31 36 130 91.2 90.0 89.0 34 40 146 23.3 27.4 100.0 Significance = 4960 1-2 Mix 0—3 Mix Total (2 SP) (3 SP) 4 3 18 10.3 7.7 12.5 35 36 126 89.7 92.3 87.5 39 39 144 27.1 27.1 100.0 Significance = 5340 251 TABLE F.2. (Continued) DEODORANT--ADVERTISING-RELATED RECALL 3-0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 7 10 9 6 32 Recall 17.1 25.6 23.1 18.2 21.1 Recall 34 29 30 27 120 82.9 74.4 76.9 81.8 78.9 Column 41 39 39 33 152 Total 27.0 25.7 25.7 21.7 100.0 Chi-Square = 1.14446 D.F. = 3 Significance = 7664 -----------—-‘-——-————--~—---‘---—-———----—--—-—---——-—--———— SOAP-~ADVERTISING-RELATED RECALL 3—0 Mix 2-1 Mix 1-2 Mix 0-3 Mix Total (3 Ads) (2 Ads) (2 SP) (3 SP) No 6 2 3 3 l4 Recall 13.6 4.7 7.7 7.3 8.4 Recall 38 41 36 38 153 86.4 95.3 92.3 92.7 91.6 Column 44 43 39 41 167 Total 26.3 25.7 23.4 24.6 100.0 Chi—Square 2.44561 D.F. Significance 4852 252 TABLE F.3. Analysis of Variance Summary Tables for Brand Value by Product Category HAIR REMOVER BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 19.994 3 6.665 .885 .458 Explained 19.994 3 6.665 .885 .458 Residual 286.292 38 7.534 Total 306.286 41 7.470 NAIL POLISH REMOVER BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 4.989 3 1.663 .392 .759 Explained 4.989 3 1.663 .392 .759 Residual 678.914 160 4.243 Total 683.902 163 4.196 MASCARA BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 2.159 3 .720 .117 .950 Explained 2.159 3 .720 .117 .950 Residual 800.863 130 6.160 Total 803.022 133 6.038 -—-——--—--——————-—----——-—----———-—-——-————-—-—o.————--——----- HEADACHE REMEDY BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 4.084 3 1.361 .279 .840 Explained 4.084 3 1.361 .279 .840 Residual 545.951 112 4.875 Total 550.034 115 4.783 TABLE F.3. (Continued) POWDERED DRINK MIX BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 23.222 3 7.741 .913 .439 Explained 23.222 3 7.741 .913 .439 Residual 601.764 71 8.476 Total 624.987 74 8.446 SANITARY NAPKIN BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 22.297 3 7.432 2.142 .101 Explained 22.297 3 7.432 2.142 .101 Residual 305.388 88 3.470 Total 327.685 91 3.601 DEODORANT BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 3.525 3 1.175 .207 .891 Explained 3.525 3 1.175 .207 .891 Residual 698.128 123 5.676 Total 701.654 126 5.569 SOAP BRAND VALUE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 16.110 3 5.370 1.226 .302 Explained 16.110 3 5.370 1.226 .302 Residual 648.101 148 4.379 Total 664.211 151 4.399 TABLE F.4. Evaluation by Product Category Analysis of Variance Summary Tables for Brand HAIR REMOVER BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 6.525 3 2.175 2.299 .093 Explained 6.525 3 2.175 2.299 .093 Residual 35.951 38 .946 Total 42.476 41 1.036 NAIL POLISH REMOVER BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 1.032 3 .344 .727 .538 Explained 1.032 3 .344 .727 .538 Residual 75.749 160 .473 Total 76.780 163 .471 MASCARA BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .692 3 .231 .296 .828 Explained .692 3 .231 .296 .828 Residual 101.137 130 .778 Total 101.828 133 .766 HEADACHE REMEDY BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .154 3 .051 .066 .978 Explained .154 3 .051 .066 .978 Residual 86.708 112 .774 Total 86.862 115 .755 255 TABLE F.4. (Continued) POWDERED DRINK MIX BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 5.229 3 1.743 1.725 .170 Explained 5.229 3 1.743 1.725 .170 Residual 71.758 71 1.011 Total 76.987 74 1.040 SANITARY NAPKIN BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .140 3 .047 .099 .960 Explained .140 3 .047 .099 .960 Residual 41.718 88 .474 Total 41.859 91 .460 DEODORANT BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .747 3 .249 .371 .774 Explained .747 3 .249 .371 .774 Residual 82.670 123 .672 Total 83.417 126 .662 SOAP BRAND EVALUATION Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .992 3 .331 .782 .506 Explained .992 3 .331 .782 .506 Residual 62.633 148 .423 Total 63.625 151 .421 TABLE F.5. 256 Analysis of Variance Summary Tables for Purchase Intent by Product Category HAIR REMOVER PURCHASE INTENT Source of Variation Main Effects Explained Residual Total Sum of Squares 22.272 22.272 362.799 385.071 3 3 38 41 NAIL POLISH REMOVER PURCHASE INTENT DF 3 3 160 163 Signif of F .335 .335 Sum of Source of Variation Squares Main Effects 22.564 Explained 22.564 Residual 1056.966 Total 1079.530 MASCARA PURCHASE INTENT Source of Variation Main Effects Explained Residual Total Sum of Squares 17.676 17.676 996.682 1014.358 HEADACHE REMEDY PURCHASE INTENT Source of Variation Main Effects Explained Residual Total Sum of Squares 8.939 8.939 929.923 938.862 Mean Square F 7.424 778 7.424 778 9.547 9.392 Mean Square F 7.521 1.139 7.521 1 139 6.606 ‘ 6.623 Mean Square F 5.892 769 5.892 769 7.667 7.627 Mean Square F 2.980 .359 2.980 .359 8.303 8.164 Signif of F .783 .783 257 TABLE F.5. (Continued) POWDERED DRINK MIX PURCHASE INTENT Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 36.578 3 12.193 1.129 .343 Explained 36.578 3 12.193 1.129 .343 Residual 766.702 71 10.799 Total 803.280 74 10.855 SANITARY NAPKIN PURCHASE INTENT Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 26.885 3 8.962 1.126 .343 Explained 26.885 3 8.962 1.126 .343 Residual 700.365 88 7.959 Total 727.250 91 7.992 DEODORANT PURCHASE INTENT Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 11.220 3 3.740 .354 .786 Explained 11.220 3 3.740 .354 .786 Residual 1298.371 123 10.556 Total 1309.591 126 10.394 SOAP PURCHASE INTENT Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 52.118 3 17.373 2.239 .086 Explained 52.118 3 17.373 2.239 .086 Residual 1148.349 148 7.759 Total 1200.467 151 7.950 258 TABLE F.6. Analysis of Variance Summary Tables for Overall Cognitive Response by Product Category HAIR REMOVER OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .632 3 .211 .428 .734 Explained .632 3 .211 .428 .734 Residual 18.701 38 .492 Total 19.333 41 .472 NAIL POLISH REMOVER OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 2.275 3 .758 1.183 .318 Explained 2.275 3 .758 1.183 .318 Residual 102.572 160 .641 Total 104.848 163 .643 MASCARA OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 1.716 3 .572 .641 .590 Explained 1.716 3 .572 .641 .590 Residual 115.986 130 .892 Total 117.701 133 .885 HEADACHE REMEDY OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .695 3 .232 .348 .791 Explained .695 3 .232 .348 .791 Residual 74.512 112 .665 Total 75.207 115 .654 TABLE F.6. (Continued) POWDERED DRINK MIX OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 1.232 3 .411 .382 .766 Explained 1.232 3 .411 .382 .766 Residual 152.850 142 1.076 Total 154.082 145 1.063 SANITARY NAPKIN OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 3.306 3 1.102 1.237 .301 Explained 3.306 3 1.102 1.237 .301 Residual 78.379 88 .891 Total 81.685 91 .898 DEODORANT OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects .879 3 .293 .588 .624 Explained .879 3 .293 .588 .624 Residual 61.310 123 .498 Total 62.189 126 .494 SOAP OVERALL COGNITIVE RESPONSE Sum of Mean Signif Source of Variation Squares DF Square F of F Main Effects 1.673 3 .558 .927 .429 Explained 1.673 3 .558 .927 .429 Residual 89.005 148 .601 Total 90.678 151 .601 B IBLIOGRAPHY BIBLIOGRAPHY Aaker, David A. ”A Measure of Brand Acceptance." Journal of Marketing Research, 9 (May 1972): 160-167. Aaker, David A. "The Long-Term Value of Temporary Price Reductions." Ph.D. dissertation, Graduate School of Business, Stanford University, August 1969. Aaker, David A. and John G. Myers. Advertising Management. Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1975. A. C. Nielsen Company. What Consumers Think of Coupons: A Study of Consumer Actions and Reactions. Clinton, Iowa: Nielsen Clearing House, A. C. Nielsen Company, 1985. Advertising Research Foundation. ”Are There Consumer Types?” New York: Advertising Research Foundation, 1964. Assael, Henry. Consumer Behevior and Marketing Action, Second edition. Boston, Massachusettes: Kent Publishing Company, 1983. Bantick, Keith. ”Sales Promotion in the Marketing Mix,” in Managing Sale! Progotiep, Julia Piper, ed. Westmead, England: Gower Publishing Company Limited, 1980, 3-11. Beem, Eugene H. and H. Jay Shaffer. Ipiggers to Ceetomer Action-—Some Elements in a Theory of Progotionel Induce— ment. Marketing Science Institute Working Paper, Report No. 81-106. Cambridge, Massachusetts: Marketing Science Institute, 1981, 1-70. Belch, George Edward. ”An Investigation of the Effects of Advertising Message Structure and Repetition upon Cogni— tive Processes Mediating Message Acceptance.” Ph.D. dissertation, University of California, Los Angeles, 1980. Belch, George E. ”The Effects of Television Commerical Repetition on Cognitive Response and Message Acceptance.” Jounral of Consumer Research, 9 (June 1982): 56-65. Belch, George E. and Michael A. Belch. "An Investigation of the Effects of Repetition on Cognitive and Affective Reactions to Humorous and Serious Television Commercials.” Advances in Consumer Research, 11 (1984): 4-10. Bem, Daryl J. "Self Perception: An Alternative Interpreta- tion of Cognitive Dissonance Phenomena." Psychological Review, 74, 3 (1967): 183-200. 261 Bettman, James R. "Memory Factors in Consumer Choice: A Review.” Journal of Marketing, 43 (Spring 1979): 37-53. Blattberg, Robert, Thomas Buesing, Peter Peacock, and Subrata Sen. ”Identifying the Deal Prone Segment.” Journal of Marketing Research, 15 (August 1978): 369-77. Blattberg, Robert 0., Gary D. 8ppen, and Joshua Lieberman. "A Theoretical and Empirical Evaluation of Price Deals for Consumer Non-Durables." Journal of Marketing, 45 (Winter 1981): 116-129. Block, Martin P. and John C. Totten. Analyzing Sales Promotion. Chicago, Illinois: Commerce Communications, Inc., forthcoming in 1986. Bowman, Russel D. Profit on the Dotted Line: Coupons and Rebates, Second Edition. Chicago: Commerce Communica- tions, Inc., 1985. Brezen, Tamara S. "Advertising Response Functions.” An unpublished doctoral preliminary examination paper, Michigan State University, April 1985. . Brezen, Tamara 8. ”Measuring Advertising Effectiveness.” An unpublished doctoral preliminary examination paper, Michigan State University, April 1985. Brown, George H. "Brand Loyalty-~Fact or Fiction?” Advertising Age. A series of articles on the following dates: (June 9, 1952): 53-55; (June 30, 1952): 45—47; (July 14, 1952): ? ; (July 28, 1952): 7; (August 11, 1952): 56-58; (September 1, 1952): 80—82; (October 6, 1952): 82-86; (December 1, 1952): 76-79; (January 25, 1953): 32-35. Brown, Robert George. ”Sales Response to Promotions and Advertising.” Journal of Advertising Research, 14, 4 (August 1974): 33-39. Cacioppo, John T. and Richard 8. Petty. "Persuasiveness of Communications is Affected by Exposure Frequency and Message Quality: A Theoretical And Empirical Analysis of Persisting Attitude Change.” Currept Issues and Research in Advertising (1980): 97-122. Cacioppo, J.T., S.G. Harkins, R.8. Petty, and D.F. Roberts. ”The Nature of Attitudes and Cognitive Responses and Their Relationships to Behavior,” in R.E. Petty, T.M. Ostrom, and T.C. Brock, eds. C it ve es onse in Zersuasion. Hillsdale, N.J.: Erlbaum, 1979. Calder, Bobby J. and Brian Sternthal. ”Television Commercial Wearout: An Information Processing View." Journal of Marketing Research, 17 (May 1980): 173-186. 262 Carman, James M. "Correlates of Brand Loyalty: Some Posi- tive Results.” Journal of Marketing Research, 7 (February 1970): 67-76. Christopher, Martin. ”Researching Below-the-Line.” Admap (December, 1971): 404-413. Clayton, Alden G. The Relationship Betweep Advertising and Promotiop: Aggge Obeervetioneg Sepculations, and Hypotheses. Marketing Science Institute Special Report, Report No. 75-110. Cambridge, Massachusetts: Marketing Science Institute, 1975, 1-29. Cotton, B. C. and Emerson M. Babb. "Consumer Response to Promotional Deals.” Journal of Marketing (July 1978): 109-113. ”Coupon Patterns: What They Mean.” Sales Promotion Mopitor 3, 2 (February 1985): 9, 11. Craig, C. Samuel, Brian Sternthal, and Clark Leavitt. ”Advertising Wearout: An Experimental Analysis.” Jourgel of MarketipgeResearch, 8 (November 1976): 365- 3 2. Craik, Fergus, and Robert S. Lockhart. ”Levels of Processing: A Framework for Memory Research.” Journal of Verbel Learning and Verbal Behavior, 11 (1972): 671— 684. Cunningham, Ross M. ”Brand Loyalty—-What, Where, How Much?" Harvard Business Review, 34 (January-February, 1956): 116-128. Cunningham, Ross M. ”Customer Loyalty to Store and Brand," Harvard Business Review, 39, 6 (November/December 1961): 127—137. QFS Progotion Report, ”A Promotion Review of 1984." New York: Dancer, Fitzgerald Sample, Inc., March 1985. Day, George S. "A Two-Dimensional Concept of Brand Loyalty.” Journal of Agvertisigg-lggggrc , 9 (September 1960): 29—36. ”Discounting Price.” Sales Progotion Mopitor, 2, 7 (July 1984): 49, 52. Dodson, Joe A., Alice M. Tybout and Brian Sternthal. ”Impact of Deals and Deal Retraction on Brand Switching.” Journal of Marketing Research 15 (February 1978): 72-81. 263 Doob, Anthony N., J. Merrill Carlsmith and Jonthan L. Freedman. ”Effect of Initial Selling Price on Subse- quent Sales.” lggrnel of Personality and Social Psychology, 11, 4 (1969): 345-350. Ernst, Otmar. ”New Evidence on How Advertising Works: Theories and Results.” Berlin Conference. Admap (February 1978): 80-87. Farley, John. "Brand Loyalty and the Economics of Informa— tion,” Journal of Busines , 37, 4 (October 1964): 370- 381. Farley, John. ”Testing a Theory of Brand Loyalty.” Proceedings of the Winter Conference of the American Marketing Association (December 1963): 308-315. Festinger, L. A Theory of Cognitive Dissopance. California: Stanford University Press, 1957. Fishbein, Martin and Icek Ajzen. Belief, Attitude, Intention end Behavior: An Introduction to Theory and Research. Reading, Massachusettes: Addison-Wesley, 1975. Frank, Ronald 8. ”Is Brand Loyalty a Useful Basis for Market Segmentation?" Journal of Agvertising Research, 7, 2 (June 1967): 27-33. Frank, Ronald and Harper Boyd, Jr. ”Are Private—Brand-Prone Grocery Customers Really Different? Journel of Adverti- sing Research, 5, 4 (December 1965): 27-35. Frank, Ronald and William Massy. ”Estimating the Effects of Short-Term Promotional Strategy in Selected Market Segments,” in Sales Progotion Analysis: Some Applica- tions of Quantitetive Iechnigues. Boston: Allyn and Bacon, 1965. Frank, Ronald and William Massy. "Marekt Segmentation and the Effectiveness of a Brand’s Price and Dealing Policies.” Journal of Business, 38, 2 (April 1965): 186-200. Frank, Ronald, William Massy, and Donald Morrison. ”The Determinates of Innovative Behavior with Respect to a Branded, Frequently Purchased Food Product.” [:ggeegings of the Winter Conference of the American Marketing Associatiop (December 1964): 312-323. Gardner, David M. ”The Distraction Hypothesis in Marketing.” Journal of Advertising Research, 10, 6 (December 1970): 25-30. 264 Giges, Nancy. ”Pepsi Counters Coke Again, Shifts from Price to Promos." Advertising Age, 51, 5 (September 15, 1980): 2, 81. Gorn, Gerald J. "The Effects of Music in Advertising on Choice Behavior: A Classical Conditioning Approach.” Journal of Merketing. 46 (Winter 1982): 94-101. Grass, Robert C. ”Satiation Effects of Advertising." Proceedings of the 14th ARF Annual Conference. New York: Advertising Research Foundation, 1968, 20—28. Grass, Robert C. and Wallace H. Wallace. ”Satiation Effects of TV Commercials." Journal of Advertising Research 9, 3 (June 1972): 3-10. Greenberg, Allan and Charles Suttoni. "Television Commercial Wearout." Journal of Advertising Research. 9, 3 (June 1973): 3—10. Greenwald, A.G. ”Cognitive Learning, Cognitive Response to Persuasion, and Attitude Change,” in A.G. Greenwald, T.C. Brock, and T.M. Ostrom, eds. Psychological Founda- tions of Attitudes. New York: Academic Press, 1968. Guest, Lester. "A Study of Brand Loyalty." Journal of Applied Psychology, 28, 1 (February 1944): 16—27. Guest, Lester. "Brand Loyalty--Twe1ve Years Later." Journal of Applied Psychology, 39, 6 (1955): 405-408. Hamm, B. Curtis, Michael Perry and Hugh F. Wynn. ”The Effect of a Free Sample on Image and Attitude.“ Journal of Advertising Research, 9,4 (December 1969): 35-37. Helson, Harry. Adaptation-Level Theory. New York: Harper and Row, 1964. Hovland, C.I., Janis, I.L., and Kelley, 8.8. Communication and Persuasion. New Haven: Yale University Press, 1953. Howard, John A. Consumer Behavior: Application of Theory. New York: McGraw—Hill Publishing Company, 1977. Jacoby, Jacob. ”A Model of Multi-Brand Loyalty.” Journal of Advertising Research, 11 (June 1971): 26. Jacoby, Jacob and David B. Kyner. "Brand Loyalty Vs. Repeat Purchasing Behavior." Journal of Marketing Research, 10 (February 1973): 2. ' 265 Jacoby, Jacob, Donald E. Speller, and Carol A. Kohn. "Brand Choice Behavior as a Function of Information Load.” Journal of Marketing Research, 11 (February 1974): 63— 69. Jacoby, Jacob, Donald E. Speller, and Carol Kohn Berning. ”Brand Choice Behavior as a Function of Information Load: Replication and Extension.” Journal of Consumer Research, 1 (June 1974): 33-42. Jarvis, L. P. ”An Empirical Investigation of Cognitive Brand Loyalty and Product Class Importance as Mediators of Consumer Brand Choice Behavior.” Unpublished manuscript, Ohio State University, 1978. Katz, 8. "0n Reopening the Question of Selectivity in Expo— sure to Mass Communications,” in R.P. Abelson, et al., eds. Theories of Cognitive Consistency: A Sourcebook. Chicago: Rand McNally, 789. Kotler, Philip. MerketinggManagement: Analysi , Plenning, and Control, Fourth edition. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1980. Lavidge, Robert J. and Gary A. Steiner. ”A Model for Predictive Measurements of Advertising Effectiveness.” Journal of Marketing, 25 (October 1961): 59-62. LoScuito, Leonard A. "Effects of Advertising Frequency and Product Usage on Recall: A Laboratory Simulation.” Proceeding of tee 76thgégerican Psychological Association Annual Conferenee, No. 75-76. (1967-68): 679—680. Luick, John F. and William Lee Ziegler. Sales Promotion and Modern Merchandising. New York: McGraw Hill, 1968, 4. Lutz, Kathy A. and Richard J. Lutz. ”Effects of Interactive Imagery on Learning: Application to Advertising." Journal of Applied Psychology, 62 (August 1977): 493— 498. Macoby, N. "Arguments, Counterarguments and Distraction,” in D.E. Payne, ed. The Obetipete Audience. Foundation for Research on Human Behavior Report. Ann Arbor: Braun Bumfield, 1965. ”Marketing Definitions: A Glossary of Marketing Terms." Chicago: American Marketing Association, 1960, 20. Messy, William F. and Ronald 8. Frank. ”Short-Term Price and Dealing Effects in Selected Market Segments.” Journal of Marketing Research, 2, 2 (May 1965): 171— 185. 266 McAlister, Leigh. "A Theory of Consumer Promotions: Mana- gerial Implications.” Sloan School of Management Preliminary Working Paper, 853—347. Cambridge, Massa— chusetts: Massachusetts Institute of Technology, 1983. Meyers, William. ”Trying to Get Out of the Discounting Box.” Adweek (November 11, 1985): 2-4. Monroe, Kent B. "Buyers’ Subjective Perceptions of Price." Journal of Marekting Research, 10 (February 1973): 70- 80. Montgomery, David B. ”Consumer Characteristics Associated With Dealing: An Empirical Example." Journal of Marketing Research, 8 (February 1971): 118—20. Nord, Walter R. and Paul Peter. ”A Behavioral Modification Perspective on Marketing.” Journal of Marekting, 44 (Spring 1980): 36-47. Ogilvy, David. Confessions of an Advertising Man. New York: Atheneum, 1964: 100-102. Olson, Jerry. ”Price as an Informational Cue: Effects on Product Evaluation," in Arch G. Woodside, Jagdish N. Sheth and Peter D. Bennetts, eds., Consumer and Indus— trial Buying. New York: North Holland, 1977, 267-286. Olson, Jerry C. and Jacob Jacoby. ”A Construct Validation Study of Brand Loyalty.” Proceedings of the 79th Ameri— can Psychological Association Convention, 1971, 657-8. Percy, Larry. "A Review of the Effect of Specific Adverti- sing Elements upon Overall Communication Response.” Current Issues and Research in Advertising: Reviews of Selected Areas (1983), 77-118. Perry, Michael, Dov Izraeli, and Arnon Perry. "Image Change as a Result of Advertising." Journal of Advertising Research, 16, 1 (July 1976): 45—50. Petty, Richard 8., Gary L. Wells and Timothy C. Brock. ”Distraction Can Enhance or Reduce Yielding to Propo- ganda: Thought Disruption Versus Effort Justification." Journal of Personality and Social Psychology. 34, 5 (1976): 874—884. Prentice, Robert M. "The CFB Approach to Advertising/Promo- tion Spending,” in The Relationship Between Advertising and Promotion in Brand Strategy, Robert A. Strang, Report No. 75-119. Cambridge, Massachusettes: Marketing Science Institute, 1975. ”Price Junkies in the Food Market." Sales Pregotion Monitor 2, 12 (December 1984): 91. staid 267 Ray, Michael L. and Alan G. Sawyer. "Repetition in Media Models: A Laboratory Technique.” Journal of egyertieing Research, 8 (February 1971): 20-29. Rethans, Arno J., John L. Swasy, and Lawrence J. Marks. ”Effects of Television Commercial Repetition, Receiver Knowledge, and Commercial Length: A Test of the Two— Factor Model.” Journal of Marketing Research, 23 (February 1986): 50-61. Roberts, D. and N. Maccoby. ”Information Processing and Persuasion: Counterarguing Behavior,” in New Models for Communication Research. Beverly Hills, California: Sage Publishing, 1973. Rockey, 8. A. and W. F. Greene. ”TV Commercial Effectiveness Under Multiple-Exposure Conditions.” Proceedings of the 24tg ARF Annual Conference. New York: Advertising Research Foundation, October 1978, 1-12. Rokeach, Milton. geliefe, AttitudesI and Values. San Francisco, California: Jossey-Bass, Inc. Publishers, 1968. Roselius, Ted. ”Consumer Rankings of Risk Reduction Methods.” Journal of Marketing, 35 (January 1971): 56-61. Rothschild, Michael L. and William C. Gaidis. ”Behavioral Learning Theory: Its Relevance to Marketing and Promotions.” Journel of Marketing, 45 (Spring 1981): 70-78. Saegert, Joel and Robert K. Young. ”Comparison of Effects of Repetition and Levels of Processing in Memory for Adver- tisements.” Agyancee for Conseger Research, 9 (1982): 431-434. Sawyer, Alan G. and Peter R. Dickson. "Psychological Perspectives on Consumer Response to Sales Promotion,” in Katherine E. Jocz, ed. Reseerch on Sales Promotion: Collected Pa ers, Report No. 84-104. Cambridge, Massachusettes: Marketing Science Institute, July 1984. Sawyer, Alan 0. ”The Effects of Repetition: Conclusions and Suggestions about Experimental Laboratory Research,” in G. David Hughes and Michael L. Ray, eds. BuyerZConsumer Inforgetion Processing. Chapel Hill, North Carolina: The University of North Carolina Press, 1974. Sawyer, Alan G. ”The Effects of Repetition of Refutational and Suportive Advertising Appeals.” Journa of Marketing Research, 10 (February 1973): 23-33. Schultz, Don 8. and William A. Robinson. Sales Promotion Management. Chicago: Crain Books, 1982. 268 Scott, Carol Ann. ”Past Behavior, Incentives, and Maintained Behavior Change: A Field Study of Self—Attribution Theory." Ph.D. dissertation, Northwestern University, June 1975. Scott, Carol A. ”The Effects of Trial and Incentives on Repeat Purchase Behavior.” Journel ef Marketing Research, 13 (August 1976): 263-269. Shimp, Terence A. ”Attitude Toward the Ad as a Mediator of Consumer Brand Choice.” Journal of Advertising, 10, 2 (1981): 9-15, 48. Silk, Alvin J. and Terry G. Vavra. ”The Influence of Adverti— sing’s Affective Qualities on Consumer Response," in G. David Hughes and Michael L. Ray, eds. BuyerZConsumer IpforgetioppProcessing. North Carolina: University of North Carolina Press, 1974, 157-186. Smith, Robert E. and Robert F. Lusch. ”How Advertising Can Position a Brand." Journal of Advertising Research, 16, 1 (February 1976): 37-43. Stevens, Nancy and Warren. ”Advertising Frequency Requirements for Older Adults.” Joppnal of Advertising Research, 23, 6 (December 1983/January 1984): 23-32. Stout, Patricia A. and John D. Leckenby. ”An Annotated Bibliography of Copy Research: 1973-1982.” Advertising Working Paper, Paper No. 16, University of Illinois, January 1984. Strang, Roger A. "Sales Promotion-—Fast Growth, Faulty Management." Harvard Business Review, 54 (July-August 1976): 115-124. Strang, Roger A. The Relegionehip Between Advertising and Promotion and Brand Strategy. Marketing Science Institute Paper. Report No. 75-119. Cambridge, Mass: Marketing Science Institute, 1975, 58-74. Strang, Roger A. and Meryl P. Gardner. ”Balancing the Communications Mix: Do Promotions Have a Negative Effect on Consumer Brand Attitudes.” Eggportegent 43 Consommateer,e§trategies de Communication, 10th Inter- national Research Seminar in Marketing, ed: J.P. Fairve and J.L. Chandon, Institut Administration des Entre- prises, Aix-en-Provence, France, 1983. Strong, Edward C. ”The Effects of Repetition in Advertising: A Field Experiment.” Ph.D. dissertation, Stanford University, 1972. ”The Bows and Whys of Coupon Behavior." Sales Progotion Monitor, 3, 1 (March 1985): 17, 19-20. 269 Totten, John C. ”Health and Beauty Aids: A Sharp Pencil is Needed.” Unpublished paper, Information Resources, Inc., Chicago, Illinois, November 15, 1985. Tuck, Robin and George Harvey. ”Do Promotions Undermine the Brand?” Adma , 8 (January 1972): 30-33. Tucker, W. T. ”The Development of Brand Loyalty.” Journal of Marketing Research, 1, 3 (August 1964): 32-35. Tulving, 8. ”Episodic and Semantic Memory,” in E. Tulving and W. Donaldson, eds., Organization of Meepry. New York: Academic Press, 1972. Tulving, E. and Z. Pearlstone. ”Availability vs. Accessabil- ity of Information in Memory for Words.” Journal of Verbal Learning and Verbal Behavior, 5 (1966): 381-391. Ulanoff, Stanley. Handbook of Sales Promotion. New York: McGraw-Hill Book Company, 1985. Vincent, Richard. ”Declining Coupon Redemption: What’s Fact; What’s Fiction.” Sales Pregotiep Mopjtor, 2, 4 (April 1984): 25-26. Webster, Frederick 8., Jr. ”The Deal Prone Consumer.” Journal of Marketing Research, 2 (May 1965): 186—189. Weinberger, Martin. ”Does the Sleeper Effect Apply to Adver- tising?” Journal of Marketipg, 25 (July 1960-October 1961): 65-67. Wright, Peter L. ”The Cognitive Processes Mediating Accep- tance of Advertising." Journal of Marketin Research, 10 (February 1973): 53-62. Wright, Peter L. "0n the Direct Monitoring of Cognitive Response to Advertising,” in G. Hughes and Michael L. Ray, eds. ConsumerZBuyer Informatiep Processing. Chapel Hill, North Carolina: University of North Carolina Press, 1974, 220-248. MICHIGAN STATE UNIV. LIBRARIES 1|HIWHIWI")“IWIHIWNWINHN”WWI 3129310880734 132'