’ - ; 1-: ~.- .. .m‘.~‘ ”.1 . d ' ' "V. »———9 Action Situation For example, in selecting a super market, hunger (or anticipation of it), and the availability of stores are equally necessary conditions for a patronage act. A super market's advertisement must be linked to the subjective interest or attitude which produced the reader's attentive response. A decision to patronize a particular super market is-a joint product of factors in the individual and factors in the situation. The super market may have placed an advertisement in the daily newspaper which served as the objective stimulus which had a correlative subjective interest or attitude on the part of the consumer, producing a "most-read" response. This response, in turn, may have served as a stimulus to building a more favorable attitude toward toward the super market, ultimately ending in a visit by the consumer. The act of patronage, then, is the crucial point in a process taking place over time. An adequate accounting scheme 48 must consider previous familiarity with the super market, as well as successive influences which may play a part in precipitating the patronage act. Figure 2. l‘ is a schematic representation of this process. Note that any present action must consider pre- vious acts. These form a sequence of stages along a time continuum. In each stage the incremental learning of favor- able attitudes is a response determined by both individual factors and situational factors at that time. The consumer (as C is not the same identical person as C l in that learning 2 has taken place, thus changing his mental entity. \ ‘g!.k 5 ‘\NA./// 5 \\\3Au/// s \\3AM// 9 \\\9A 1 _1_ 2 2. 3 .2. 4 .1 sees ad friend needs store buys recommends food available Fig.2J.Representation of Evolutionary Analysis of Action Beginning at the left with the consumer (C ) at the time she reads an advertisement (5 ) for a particular super market, her response (A’) is a favorable feeling toward the store. This changed attitude leaves her a different person (C ). A friend recommends the store (5 ), increasing her knowledge and favorable attitude toward the store (A ). Now an even further changed person (C ), she discovers she must buy food for her next meal (3 ) and makes the decision (A ) to go shop- ping for groaeries. The patronage act ( 4) occurs when the consumer (now C ), with an attitude favor- able to the store and for making a food purchase, finds the store open (s4). The time line above may 49 be only a few hours in duration; however, for practical purposes, an inquiry should begin at the stage where the want and means for gratifying clearly exist in the person's mind. It can be assumed the housewife already had a well-defined complex of beliefs, atti- tudes, knowledges, and needs related to her food store'behavior. ~By using the panel technique, it is possible to analyze at least two stages of the patronage process by repetitive interviews with the same respondents. Appropri- ate instruments can then enable the observer to analyze the relationship between selected situational and personality factors and a consumer's super market patronage over the period of time encompassed. Figure 2.2 incorporates variables and constructs which are relevant to super market choice. It will be noted that the diagram is atomistic, viewing the behavior of one consumer making a single choice among a narrowly defined set of alternative super markets, and only partly dynamic in that any action must be viewed on a time con- tinuum - consequences of previous acts affect attitudes which in turn relate to the current decision - therefore, the effects of such changes are not revealed. As will be seen in Chapter III, the study was designed so as to hold constant many of the situational forces surrounding super market attributes. The selection 50 pomwmw GOAumoHCDEEoo msaumamsm How Emflomumm m mm comm Doolommsouumm pomeEnmmsm I N.N oNHsouumm Dos unecoupmm Acoauamommuov ZOHmHUmQ QWWflAMQ S mNHsouHmm #0: M mNHgouumm U Anouuomv _R\\m onmHome meaHomst ATT (learning-meaning) Q .U .m .¢ OHOum mUHosU muoum Bummmm mmsa>mm\wEoosH mnouomm UHEocoom oa0u\m5umum coepmmsuoo coflpmosom wHOpomm Hmfloom mom .usmamz .usmflm: mnsuosuum-amoam>£m mMHuHHHQm smasumse meuHHHQm huomcom Nww mnouomm HMUHUOHowmmam mmmcmsoHHmmmum .mamom .mommc msumpumm venue moosufluum .mmoflamn ousomaaaouca mmomazosx muswflnmmxm HOHHQ mosHm> rmHouUmm HMUHmoHonohmm mMOEUfim mmSDmZOU PERCEPTION (Interpretation) A.Uuo .mm>flanoH omfl'm .mosoanm Eoum wofl>omv mxuo3umc Hmcomnmmuumqfl muHEnomsou Mom mmHSmmon mayo: msoum muouomm pngGOHH>sm ameoom masmcmamoamm Hmcomuom mom mo psospmouu mom mo mQHEHu \\ mom mo mNHm ocm HmQESs A now no usmucoo poms mHooE msamauuo>om mu0pomm mmmmmoz wuoum flouo .mmfluflaflomm mcwxumm mwua>uom stcomuwm >pmanm> omaosmsuume mmmsflasmmao GOHHMUOH moanm mmusuxum a mcfloaflsn huflamsv mmaogmnoume Q .U .m .m wuoum mmpdnfluupfi wnoum mm080 23.8% Neighbor, other 13.2 9.5 Friend, school or job 9.0 4.7 Relative 4.7 5.8 Other 2.6 0.5 None 18.0 55.0 TABLE 4.30 SUPER MARKETS DISCUSSED "MOST" BY PANEL "LAST WEEK" (OCTOBER) Store Per Cent A & P 20°6 Note: When categorized by Kroger 12.7 . store customers, roughly National 11.1 . one-half of the percentages Wrigley 10.1 shown for each store were Other 15.8 composed of customers of None 18°C that store Don't remember 11.7 ' Summarizing the other questions concerning face- to-face communication patterns, it was found that 48 per cent of the panel had talked to someone prior to making a decision to patronize their preferred store. Twenty per cent of the panel stated it was the most important influence in making their decision. Of those talked to, neighbors accounted for over half of this influence. 129 Further analyses will be made of the interpersonal patterns in Chapter VI. ANALYSIS OF CHANGES IN STORE PREFERENCE Tables 4.31 and 4.32 show the pattern of change within the sample between October and January relative to store patronage. Of the original 189 members of the panel, twenty moved away, while forty-four shifted their preference to another store. The distribution of the 26 per cent who shifted allegiance shows the A & P the major loser with four- teen customers going elsewhere, or nearly 25 per cent of those preferring in October, while gaining only seven new 'Villagers (or 12 per cent). In two previous surveys, the .A & P had maintained a constant percentage, therefore the decrease is meaningful in terms of the attitudinal profiles shown in Chapter V. Kroger was the significant winner during the interim, losing six customers and gaining sixteen, up 25 per cent. There was little difference in the gains and losses of the remaining stores. The pattern of shifts is note— worthy, the A & P shifters scattering to all other stores, with National tending to be the largest gainer. National shifters resembled the A & P group. Wrigley shifters nearly all changed to Kroger. Kroger shifters were divided between the A & P and Wrigley. As will be seen in Chapter V, the 130 image analysis predicted that this would be the probable pattern of the changers. There was no significant difference between old and new residents in the proportion of changers. Of the seventy- eight new residents, 23 per cent changed, while 27 per cent of the ninety-one older residents changed. Of the forty-four changers, 42 per cent were new residents and 50 per cent were old residents. No difference was found between the changers and non-changers on any of demographic or socio-economic variables. TABLE 4.31 CHANGES IN STORE PREFERENCE OCCURRING IN PANEL BETWEEN OCTOBER, 1960 (n=189) AND JANUARY, 1961 (n=l69) Direction of Shift Store , , Moved Lost to Other Gained from Net Gain Away Stores Other Stores or Loss A & P 4 14 7 -11 'Wrigley 6 8 8 — 6 Kroger 5 6 16 + 4 National 3 9 6 - 6 Prince Bros. 1 3 3 - l Schmidt's 0 2 3 + 1 Other 1 2 l - 2 Total 20 44 44 -20 131 TABLE 4.32 DISTRIBUTION OF CHANGERS IN PANEL, JANUARY, 1961 (n=44) First Choice in January 1961 First Choice in October S E u w H c m p 1960 a. H m o o -H u H m m -H c E m m w -H o u -a a a .u H H m H U u o a: 3 M z m m 0 E4 A & P 2 3 5 l 2 l 14 Wrigley l 6 1 I 8 Kroger 3 3 6 National 2 2 3 1 1 9 Prince Bros. 1 1 l 3 Schmidt's l l 2 Other 2 2 Total 7 8 l6 6 3 3 1 44 Analytical Framework for Analyzing_flChanqers.” It was assumed that an unknown portion of the panel would change their first choice super market during the interval between the first and second waves. Chapters V and VI will describe the findings relative to attitudinal changes occurring in the panel, including the "changers." Findings from the pilot study, however, indicated that certain additional information would be helpful in order to provide a minimum account of a change in store patronage. Accordingly, the reinterview 132 questionnaire contained several questions (see Wave II Questionnaire, Appendix, pp. 7.8) which provided data rela— tive to three elements felt to be essential: l. The "changer" must be given the opportunity to answer in terms of the new choice. In addition, such factors as satisfaction—dissatisfaction with old store are relevant, though they would not account for the direction of the new choice. 2. Questions for assessing each of the communications influences, together with specific information questions for categorizing reasons, if any, why the communications were effective. The channel and message are neither mutually exclusive nor both necessary for a complete response. Ideally, however, answers which can provide information concerning message content as well as channel pressure assure a complete response. Probe-type questions were used on the assumption that respon- dents would tend to feel reasons for changing were self-evident and would tend to give superficial answers, e.g., "I felt Kroger did not charge as much." 3. The type of change had to be derived in order to find out if the respondent had learned new facts about the store she now preferred, or whether her attitudes had changed so that the new preference fitted them better. These elements are shown graphically in the diagram below: Old Choice (change due to New Choice (What attributes dissatisfaction or greater did store have interest in New) that changer felt \\\\\\\A to be desirable?) Communications Channel (Where did changer learn about considerations leading to change?) Channel Incidents (What was said, read, and in what direction?) Revaluation of choices (What caused the change?) 133 After an interviewer had determined that a respon— dent had shifted her store preference, he asked the question, "When you were interviewed in October, you stated that (former super market) was your first choice. Why did you stop going to (former choice)?" Interviewers were instructed to probe to find out if a specific incident might have triggered the switch, such as a refusal to cash a check. It was already known from the pilot study and Wave I that respondents would tend to answer in terms of store attributes (low price, quality of meats, quality of produce, etc.) to such a general question. The question also served to stimulate conversation about the change, encouraging better recall for the later questions. In eliciting information about other influences playing a part in the switch, experience had warned us not to suggest in any way that the "reasons" given by a respondent were not valid so as not to incur her displeasure. Rightly or wrongly, the respondents seemed to firmly believe her choice had the "best" prices, meats, etc., a phenomenon which might be referred to as "super market rationalization." Followingthe probe‘question, a more directed question was asked to assess the role of satisfaction-dissatisfaction in the change process. It was hypothesized that two types of changers would emerge: those who were "pushed" away from old store because of unfavorable experience, and those who changed 134 for other reasons although satisfied with old store, or "pulled" away from old store. The question was asked, "Would you say, in general, that you were satisfied or dissatisfied with the (old store) at the time you decided to change?" The respon- dent was asked to check a seven-step scale ranging from "very satisfied" to "very dissatisfied." To find out if any specific "pull" occurred relating to the new choice, a probe-type question was asked, "How did you happen to decide to patronize new store?" Actual channel influence questions were then given, the first dealing with family influence ("Did anyone in your family say anything about (new store) before you changed?"), and repeated for the other face-to-face and the various adver- tising channels. Finally, the respondent was queried as to the relative importance of the channel influences so as to distinguish between those which were contributory and those which were effective: "Summing up, now, what was the most important source of information which helped you to decide upon the (new store)?" In review,the changer questions attempted to (1) get an indication of the role of satisfactionedissatisfaction; (2) find out if a store—related incident might have occurred to "cause" change; (3) find out if respondent was exposed to face-to-face and mass communications influences; (4) assess 135 which influence was felt by the respondent to be a decisive factor. The result of the exposure and assessment questions enabled an impact rating to be made for the group which could be construed to be a one measure of effectiveness for the message influences. Findings. Table 4.33 reveals a completely unexpected finding. Only 25 per cent of the changers stated they were dissatisfied with their former preference while 42 per cent actually expressed a high degree of satisfaction. The sugges- tion is that the majority of changers do so because of "pull" to the new store rather than a "push" away from their former choice. TABLE 4.33 SATISFACTION INDEX OF CHANGERS WITH "OLD" STORE AT TIME OF CHANGE (n=44) Index of Satisfaction Percentage Very satisfied 11.6 Quite satisfied 30.2 Slightly satisfied 16.3 No feeling 16.3 Slightly dissatisfied 4.6 Quite dissatisfied 9.3 Very dissatisfied 4.6 No response 7.0 136 Table 4.34 shows the reasons given by the changers, for changing stores, classified into ten categories from the responses to the open-end question. The reasons given for leaving the store formerly preferred are almost the same as the responses given by the entire panel relative to the fac- tors considered most important in selecting a super market (see Table 4.12). However, when they discussed the reasons they preferred the new store, an entirely new pattern emerges. Price is.no longer seen as being of paramount importance. The mention of quality of meats and produce also drops dramati- cally. Convenience remains the same. Instead of store attributes, the changers now indicate the effects of communi- cations influences on their decision, face-to-face influence being the most important followed by advertising. While it can be argued that the communications influences concerned store factors, it must be remembered that the major stores were nearly equivalent choices in fact, and that the changers were generally satisfied before switching preferences. It seems reasonable, therefore, to conjecture that the communi- cation influences themselves influenced attitudes precipita- ing change among many of the changers. 137 TABLE 4.34 PERCENTAGE OF CHANGERS ACCORDING TO STATED REASONS FOR CHANGING PREFERENCE (n=44) Reason for Reason for Factor Leaving Old Preferring Store New Store Price 18.6 9.3 Quality of meats/produce 18.6 7.0 Convenience 14.0 14.0 Face-to-face talk 14.0 23.2 Stamps 11.6 16.3 Advertising 7.0 14.0 Specific incident 7.0 --—- Other reasons 6.9 6.9 Cleanliness/attractiveness 2.3 0.0 Formerly patronized store ---- 9.3 One event occurred during the interim between the two waves which may have contributed to the emergence of the Kroger store as the principal winner in the competition for customers. The A & P began advertising in the State News, the campus daily, shortly after Wave I had been completed. The ads were usually about forty column inches in size and appeared once each week. In November, the Kroger Company began a series of full page ads, also on a once—a—week basis. Samples of the advertisements by the two stores are included in the Appendix. In the judgment of members of the Depart- ment of Advertising, MSU, the Kroger advertisements were clearly superior to the A & P advertisements. (The A & P ads were prepared by the local store manager, the Kroger ads 138 were prepared by the advertising manager in the KrOger divisional office.) As neither store had inserted an ad in the State N§w§_prior to the readership study on Wave I, no readership results were obtained. Therefore, approximately 10 per cent of the panel were contacted during the last week of November and shown copies of advertisements by the two stores. The Kroger advertisements were selected as "best" by all but one person. The implication is that the A & P may have actually suffered by their very act of advertising in the State News. Table 4.35 indicates the percentage of changers who could remember seeing specific advertisements for their new store choice prior to changing their preference. Of the 25 per cent of changers who "saw" ads in the State News, almost four-fifths were respondents Who changed to Kroger. Table 4.36 shows the relative importance accorded to the various media by the changers. Here, direct mail is revealed as the second most important medium, but newspaper advertising leads by a wide margin. Table 4.37 shows the pattern of interpersonal influence. The influence of neighbors is not as strong as expected, the changers showing no significant differences between the four types. Interpersonal influence acting as a source of social pressure and support, as well as acting TABLE 4.35 PERCENTAGE OF CHANGERS WHO SAW NEWSPAPER ADS FOR CURRENT FIRST CHOICE SUPER MARKET BEFORE PATRONIZING IT (n=44) Newspaper Percentage Saw no ads 37.2 Lansing State Journal 30.2 Michigan State News 13.9 Saw ads in both papers 11.6 No response 7.0 100.0 TABLE 4.36 CHANGERS' RATINGS OF MASS MEDIA WHICH ADVERTISED CURRENT FIRST CHOICE SUPER MARKET BEFORE CHANGER PATRONIZED IT (n=44) Rating of Source of Information Media Newspaper Radio TV Direct Mail Important Recalled specific ad 39.5 4.7 7.0 18.6 Didn't recall ad 13.9 4.7 4.7 27.9 Not Important Recalled specific ad 4.7 9.3 7.0 20.9 Didn't recall ad 34.9 74.4 74.4 25.6 No response 7.0 7.0 7.0 7.0 100.0 100.0 100.0 100.0 139 140 TABLE 4.37 CHANGERS' RATINGS OF TYPES OF FACE-TO-FACE SOURCES (n=44) — fi f Source of Information . . 0 Rating Of Neighbor, Neighbor, Friend, ther Source Not (Relatives, Same Floor Other . Neighbor etc.) Important Talked 18.6 13.9 11.6 18.6 Didn't talk 16.3 11.6 16.3 30.3 Not Important Talked 11.6 9.3 7.0 0.0 Didn't talk 46.5 58.1 58.1 44.2 No Response 7.0 7.0 7.0 7.0 100.0 100.0 100.0 100.0 as a channel, can be seen in the responses of those rating other people as "important," but had not talked to anyone. Even without direct conversation interpersonal influence is apparent. The tendency to follow others can be seen, for example, in the 30 per cent who considered relatives important but did not talk (Pmother used to shop there.") Table 4.38 reveals the relative rankings accorded three message types: in first place is predisposition; second is face-to—face; and third is advertising. All were shown to be important factors in the decision process. 141 TABLE 4.38 PERCENTAGE RANKING OF INFORMATION SOURCES STATED BY CHANGERS AS INFLUENCING THEM TO PATRONIZE CURRENT FIRST CHOICE SUPER MARKET (n=44) Source of Information Ranking Looking for Self Face-to-face Advertising Most important 37.2 32.6 23.3 Second 39.5 23.3 18.6 Third 14.0 25.6 34.9 No response 9.3 18.6 23.3 100.0 100.0 100.0 Group Influence. Face-to-face communication has been revealed as an important factor in super market choice. However, as Katz and Lazarsfeld (1955, Part I) point out, the influence one person may exert on another not only depends upon the relationship between the two, but also upon their relationship to the primary groups to which they belong. Primary groups can be classified as (1) the informal group- ings, such as the family, neighbors, and other friends, and (2) more formal groups, such as club and church organizations. "Membership" implies internalization to some degree of the group norms and values by the individual. While this study was not designed to investigate the influence of reference groups in a systematic manner, it 142 was felt that the residents living on the same floor of an apartment block would tend to form friendships with one another. The informal groups thus formed would then develop norms and expectations concerning food store choice, a topic of common interest. The importance of group influence as a variable intervening between the mass media and the masses to modify message effect has been "rediscovered" only recently (Lazars- felg, §§_§l, (1948) Katz and Lazarsfeld, 1955; Rogers and Beal, 1958; Festinger, g§_§l., 1950). There have been few studies in the marketing-adver- tising area which have attempted to assess group influence. Whyte (1954) found that home air conditioners were clustered rather than randomly distributed in homogeneous neighborhoods, although he found no evidence that the influence extended to the brand purchased. Glock (1955) found significant clusterings of smokers and non-smokers within work and friendship groups, evidencing social pressures. Fisk (1956) found reference groups in food buying are based on kinship more often than propinquity. The major methodological problem was how to take account of interpersonal relations and still preserve the economy and representativeness of the random, cross-sectional sample. By defining the reference group to be studied as those residents living on the same floor of an apartment 143 dwelling, the panel was designed to include fifty-five floors of the 236 floors comprising Spartan Village complex. Every new and every other old resident was interviewed on each floor. By comparing store patronage patterns by floors at the begin- ning and the end of the study period, some evidence of group influence might be obtained. Assuming new residents to be more likely to be influenced by such pressure, there would be a bias in favor of a positive finding due to the fact that nearly half of the panel were newly-arrived. Analysis of Group Influence on Changers. Due to the complex probabilities involved in comparing the two waves, a second check on group influence was designed, using enumer- ation statistics, to account for the forty-three changers found in Wave II. Each family in the panel was assigned an arbitrary weight, based on position in an "agreement pattern." The highest weight was accorded to patterns which had the lowest probability of occurring, e.g., there were eight floors on which four families were interviewed and where three of the four families chose the same store. In Table 4.39, this is shown as a "3-1" agreement pattern. The eight floors in- cluded thirty-two families, twenty-four of which are shown as being "with majority" and eight as "not with majority." 144 TABLE 4.39 POSITION OF FAMILIES IN AGREEMENT PATTERNS, OCTOBER, 1960 (n - 188) Agreement Patterns Family Position With Not with Number of With Not with Majority Majority Floors Majority Majority 3 l 8 24 8 3 1-1 1 3 2 2 O 5 10 2 1 5 10 5 2-2 0 2 8 2-2 1 2 _§. 2 63 2-1-1 9 36 2-1-1-1 2 10 1-1 2 4 1-1-1 14 42 l-l—l-l 4 16 125 The question was asked: is the passage of time or the influence of floor groups effective in raising or lowering the proportion of residents patronizing the same stores on their respective floors? Inasmuch as the changers completely describe the movement between the "with" and "not with" categories between waves, the following contingency table could be constructed: (Note: Only 43 of the 44 families who changed their store preference are shown, one family was omitted.) 145 January October With Not with Total Majority Majority With Majority 3 5 8 th with Majority ;9_ .25 .35 Total 13 30 43 Because the proportions are not independent, the null hypothesis that the difference between the two propor- tions was zero was tested by using the test for correlated pro- portions. The increase in the proportion patronizing the same stores was significant at the .05 level. The hypothesis that this particular type of a reference group would influence changes in super market preference was therefore accepted. The conclusion follows that, for Spartan Villagers, group influence arising from contiguous dwellings within apartment blocks is positively related to changes in super market preferences. While this arbitrarily defined reference group was thus shown to be a factor in patronage changes, its impor- tance was not clearly established in view of the small number of respondents involved; the net gain of five in the "with majority" category barely met the level of signi- ficance. It suggests, however, that had we accounted for other reference groups, group influence would have been shown to have played an even greater role in influencing 146 super market patronage during the three-month period. (As shown in Table 4.37, many changers reported that other neigh- bors, friends, work associates, etc., were important sources of information about super markets. This study, however, was not designed to investigate these relationships in a systematic manner. The finding relative to floor influence thus agrees with the considerable evidence supporting the existence of group influence. Neighborhood groupings may give rise to "group norms" about super markets which will influence the patronage of some of its members. SUMMARY The objective of this chapter was to set forth a description of the population studied in terms of their socio-demographic characteristics, communications and store behavior patterns. Validation by further research into a more extended environment is needed. Yet many of the above findings suggest profitable application by the practitioner. The analysis reveals that the consumer tends to respond in accordance with a standard pattern of "accept- able" responses when asked to give reasons for choosing a particular store. As will be seen in the next chapter, these reasons seem to reflect social norms and values which are-selectively perceived in the store people prefer among alternative st the aggregate, a super market price, quality and variety of factors by sto rankings may r causes for sel Fact to the immedia omic object su social and fun and preserving the object. Revi shopping behav ed to: (1) (2) (3) (4) (5) (6) 147 ores which are objectively comparable. In factors stated as most important in selecting resembled those reported by similar studies: of meats and produce, cleanliness, location, merchandise. However, analysis of these re customer groups over time indicated such esult from patronage and are not primarily ection. ors for store choice should not be limited te situational factors surrounding an econ— ch as a super market, but extended to other ctional influences which operate in shaping the beliefs and attitudes people have toward ewing the findings concerning patterns of ior patterns, the members of the panel tend- shop alone or with members of the family. not with others. choose a super market previously patronized when moving to a new area. shop once a week for food needs, do all their shopping in one store. make super market shopping a principal activity which is seldom done in connection with other shopping activities. investigate alternative super markets within a short time after arriving in the community. establish patterns for preference that remained stable for the majority of the residents. (7) (8) (9) (10) 148 select among super markets according to socio— economic status. select among super markets according to age categories. in 91.6 per cent of the families, the wife decided which super market to patronize. patronize own store each week; only one in five spend more at another store. From the findings on communication patterns: (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) new residents tended to read newspapers more than old. respondents in the older age categories and who had children tended to read newspapers more than those who were in younger age cate- gories without children. less than half of the State Journal readers saw super market advertising in more than two of the four editions carry advertising "last week." of the sixteen super market ads appearing "last week," 13 per cent of thosereading the newspaper had not noted any ads, oneethird had not "read most" of any,ad, and two-thirds had not read the "coop" portion of agy ad. the same small group (about 10 per cent) of readers appeared to account for a large pro— portion of "read most" and "coop" readership. neither the position of an ad in a newspaper nor the day of the week it appeared seemed to affect the readership scores. respondents tended to read the ads for their chosen store more than ads of other stores. Readership scores by Starch categories for "own" store ad revealed a 50:40:30 per cent pattern, but varied widely for ads of alter- native stores. Kroger had the highest readership of any ads. judged to be a function of their superior treatment. direct mail was revealed as an effective mass per cent medium, at least on a short term basis, by the 10 per cent net gain in patronage "last week" by two stores using it. most respondents talk about their chosen super market, usually to friends and neighbors. (ll) (12) 149 face-to-face communication patterns revealed that each major super market was talked about "last week" by 10-20 per cent of the panel. overall, each of the four major media and the face-to-face channel had nearly the same number of people in their respective audiences, about 80 per cent of the panel, "last week." Relative to "changers," the findings indicated: (1) (2) (3) (4) (5) (6) (7) (8) about 25 per cent of the panel changed prefer- ence in a three-month period. there was no difference between changers and non-changers according to demographic and socio- economic characteristics, or length or resi- dence in the community. changers tended to be satisfied with previously preferred super market at time of change, and "pulled" to a new choice. the reasons stated for leaving a store resembled "conventional" reasons usually given for selec- ting a super market. interpersonal influence was rated as important by 30 per cent of the panel even when verbal contact was not reported. 37 per cent of the changers rated "self" as most important source of information. 33 per cent said "face-to-face sources, and 23 per cent stated advertising was most important. 53 per cent of the changers rated newspapers an important source of information concerning super markets, 19 per cent stated direct mail was important, while radio and television were rated as important by less than 10 per cent. the influence of reference groups, defined as the families living on the same floor of an apartment dwelling, was positively related to changes in super market preference. 150 CHAPTER V ,SUPER MARKET IMAGES The data from the semantic differential provided an opportunity to investigate two attributes of the measur- ing instrument: (1) its ability to provide graphic images which reveal both the quality and intensity of the feelings consumers hold toward an economic object such as a super market, and (2) the use of this data for clustering the underlying images (profiles of stores) into a small number of representative variables, or types. The pilot study revealed both could generate valuable insights for guiding management in taking reinforcing or remedial action. The central concept is that of "image,f which was discussed in Chapter II. The first section will detail the methodology and findings on the graphic images. In the second section, the typologies resulting from applying linkage analysis will be presented. The methodology and findings are discussed relative to their theoretical and practical implications in the summary. GRAPHIC IMAGES In this section, two ways of graphically describing the quality and intensity of the attitudes held by the panel toward the super markets are investigated: (1) in profile form, and (2) in three-dimensional form. 151 The data for constructing the graphic images were obtained from the semantic differential described in Chapter III. From a theoretical viewpoint, it was assumed that the connotative images of the super markets are constantly being modified over time due to the impact of communications about them on the consumer: (1) face—to-face contact with others, (2) communication from the mass media (primarily advertising), and (3) self-messages arising from predisposi- tions, direct experience, or personality changes. Several systems of comparison will be shown: A. Profiles 1. Comparisons of the various store images by the entire panel at a given time and between points in time. 2. Comparisons by panel sub-groups of the various store images held by the groups at a given time and between points in time. B. Three-dimensional Analysis 1. Comparisons of the various store images held by customer groups, using three scales, at a given time. 2. Comparisons of the various store images held by customer groups, using three scales, at different points in time. The graphic images shown in Figs. 5.1 - 5.23 fall into two categories. Shown first (5.1 - 5.15) are the store images in the form of profiles. Second, the data is pre— sented in three-dimensional form. 1523:“ Procedure for Constructing Profilgs. The nine scales on which each store concept was evaluated were equal- interval ordinal scales. Each of the seven intervals was weighted by assigning a weight of one to seven. Individual responses were entered onto IBM cards. These in turn were converted into individual and group mean scores for each scale by computing the cumulative total of frequencies times inter- val weights divided by the total number of responses. The group means were than plotted in graphic form to yield profiles of each store concept. For the various profiles, when ng and sizes of variance are taken into consideration, means which are .5 or more of an interval from the neutral (4) position at the center of the scale can be considered significantly greater than 4 at the .05 level for most scales. Standard errors of the means ranged from about .2 to .4 or an interval. By' this criterion it can be said the concepts touch off con- notative mediating responses indicating strong evaluative meanings, or attitudes (McNelly, 1961, p. 29). Single Item Evaluation. Differences between means for store couplets along the adjectival scale can also be analyzed. For example, in Fig. 5.1, the difference between the means on the clean/dirty scale for the A & P and Wrigley are significant at the .05 level (t = 1.85 > 5.05 = 1.66, df mméfific $21— mmxtJU 233* «F B 1)- d)- dP J PA >. “6...) “if...“CHCP 1 J m >32 :DoxiL J '1)- di- m>~r241vmd MBLWZNQXMZH F b L 1 q 4 74¢. 55.: dmOJU p F ii» A 4 d .rw—CZQHxEvHD >JQZHsz P p L 1 q J p .r l E: + 493th 1 d . u r 3%.?14ird. p ,a.z«.m3 nuczai CO 9:20.21 In LPN. .mfm _ @930 so.— .1930 $03? p P p p .i ‘ d 1‘) d d M>HF<>deZOUp h - ”mama moommb T I q d C P M>~mZMQXH p p - ”Nu/HmZMnmXHZHL 1 d I d J .HZ < .HWHQ m mOJ U P P bi L 1 a d . WJCZMHZLHZD MHHQZNHmrmL P D r 1 a u d > .. ~32 Z2 nwcCcp .«c .n.o:.«o.~& u: M. .m 31er mm. (do so; .53 u 10...: T i 4 I4. 4 u 1. VI >4 $6.... Zwmfrr: .89“ n b . p “wk/Hmmfl 200.1? 1 q _. d u ‘. ) .7, . a .Y .. .. ) -. _. , . .4 P ;N..~.f»~ w *dt. .w P P .1 PM P m.rH.hH/.rwawym.tr 1 d d -, .1 d i- s . .. f}. P _ .x a. h _ b b b AWOJUL 1 d d d d P >Jauwm..~_1n~v.3 F b r . Viamggdmxkb q d d 4 d . a w r>,_m~,~ p P b .’ . b > 4 J 46477.; .../..2 I L < _.._ .r «X . ..1 95 . do "1‘ JP 4» L. w GOA — N. J D V M .N ~ .32 Susanna... .3 w>m>> so 3ch spice >0 mwcfidn Co 9330an .... m .m .mfm 156 = 95). On the other hand, Kroger and National show no significant difference on the inexpensive / expensive scale (t = .59 < t .05 = 1.66, df = 95). Why did the Wrigley store depart significantly on the clean / dirty scale as compared to the other stores on the pilot study? Several respondents said it was a "messy" store. Professional super market people reported the check— out counters were poorly arranged and that stocking of shelves during busy hours caused cluttered aisles. Dark walls and a low ceiling contributed to a feeling of "dirty" in the sense of not being "clean-bright-white." More definitive reasons could conceivably have been obtained through depth interviews. In the case of Wrigley, it happened that the store was remodeled following the pilot study, eliminating most of the objectionable physical features noted above. It is interesting to note that the Wrigley profile for January (Fig. 5.2) shows a nonsignificant variation from the means of the other stores. The Wrigley mean was 3.8 in the pretest, and 3.0 in the January 1961, wave. The Kroger super market, ranked third in overall preference in the pilot study, rose to first place among the stores at the time of the January, 1961, wave. The profile data show the evolution of its increasing popularity (Figs. 5.1, 5.2, and 5.3). It is noteworthy that the price 157 scale remained almost exactly the same throughout, while there was an increase in favorability toward "fair,” "pleasant," and "friendly." Theme of Kroger advertising messages stressed "let's go Krogering——the happy way to shopl" In Wave II, Kroger clearly stands out on these scales, sug- gesting that customers were moved in the desired direction by non—price appeals. Control Concept. To provide a realistic basis for evaluating direction and intensity of attitudes other than by inter-store comparisons, ratings were obtained on a seventh concept, "My Ideal Store." The profiles for the control concept show the consumer to have decided expectations of her super market. While more extreme than the competitive store concepts on the clearly evaluative scales, as might be expected, it was not merely a matter of the respondent checking the extremes on the scales. The nine pairs of polar adjectives selected following the pilot study included four pairs from the classic list of fifty word pairs factor—analyzed by Osgood (1957, Chap. 2) having heavy factor loadings on the evalua- tive dimension. The remaining five pairs (expensive / inexpensive, friendly / unfriendly, progressive / conserva- tive, close / distant, high class / low class) were included because they were felt to be meaningful in the context of 158 super markets and thus more appropriate to the problem. For example, the reaction to progressive / con- Servative was not in the anticipated direction. It was assumed that the often-observed tendency of Americans to prefer the "new and different," referred to by the economist as "innovistic competition" (Barnet, 1956) and by the socio- logist as a major value-orientation (Williams, 1956), would be strong in the case of super markets. The profiles in- dicate that this generalization should be interpreted with caution as a factor in the selection of super markets. Schmidt's judged by an independent jury of super market executives to be the most "modern" of the stores, was also seen by the panel as the most progressive. However, this factor was a deterrent for many potential customers among the respondents, a frequent comment being that Schmidt's was ex- pensive (the store was actually among the lowest in prices). Further evidence of the utility of the control concept can be seen by inspecting the profiles on the clean / dirty scale. This scale consistently received the highest rating from the respondents, showing cleanliness to be of prime importance. The only exception can be seen in Fig. 5.10, where the customers of the A & P (the store suffering the greatest customer loss during the study period) tended to place less stress upon this factor. It will be recalled that the 159 customers of the National store (Table 4.14) chose clean— liness as the most important factor for patronizing this store. The attitudinal profiles for the National customers also revealed this group to have the highest mean scores on the clean / dirty dimension for both the ideal and own store concepts. National, the newest of the stores, apparently was doing the best job in meeting the criteria many women set forth under "cleanliness." (For the panel as a whole, women with children tended to select "cleanliness" as an important factor in store choice to a much greater degree than women without children. One might speculate that there is a keener awareness of the need for cleanliness among young mothers arising from the current norms surrounding baby care.) The findings on shopper preference for cleanliness in super markets supports findings by the Jewel Food Stores in Chicago, who now actively promote their stores as "clean and white" (Martineau, 1957, p. 179). Using Fig. 5.2 (mean scores for entire panel of 189 respondents on Wave I) as a basis for describing the "Ideal” profile, five of the adjectival scales (good / bad, fair/ un- fair, pleasant / unpleasant, clean/ dirty, friendly / unfriend- ly) have means ranging from 1.4 to 2.0. Price (expensive / inexpensive), 2.6, and location convenience (close / distant), 160 at 2.4 were not as strong as anticipated. Variability about the means showed the greatest dispersion for expensive / inexpensive, somewhat less for progressive / conservative, and low class / high class. The remainder of the scales showed very little variability. Profile findings for Panel. Figures 5.1 - 5.3 are the profiles obtained for the six super markets on three separate occasions. Fig. 5.1 reveals the profiles of mean scores for the sample used in the pilot study in January, 1960. They are included inasmuch as the data were obtained from a completely different sample and at an earlier time, thus giving a richer basis for evaluating the semantic differential against the usual criteria for measuring in— struments: objectivity, reliability, validity, sensitivity, comparability, and utility. The profiles present a readily discernible patterning of attitudes. The reliability of the instrument can be noted in the very similar profiles in Figs. 5.1 - 5.3, despite the time difference for all three and the different sets of respondents studied (Fig. 5.1 versus 5.2 - 5.3). The validity of the instrument is evidenced by close / distant scale. The shopping center stores were located approximately two miles from Spartan Village, the A & P was three miles, while Schmidt's was four miles distant. The “.161 means for the shopping center stores are almost identical, the A & P and Schmidt means deviate almost in proportion (2:3:4) to the actual physical distance. The difference in the mean scores for the "Ideal" concept in Fig. 5.1 as compared with the Wave I and II profiles in Figs. 5.2—-5.3 can be accounted for by the fact that the pilot study in— cluded respondents living in another married housing develop— ment located a half—mile closer to the shopping center, they tended to regard location convenience as more desirable in their "ideal" scores. Likewise, the mean scores for A & P and Schmidt's show a greater dispersion in the pilot study profiles, but are proportionally similar to the panel profiles. Profiles of Changers and Non-changers. Figs. 5.4 through 5.7 reveal the profiles for changers and non—changers on Waves I and II. The profiles for Wave I were obtained by taking the changers and non—changers, as ascertained in Wave II, and extracting their IBM data cards from Wave I for analysis. The profiles for the non-changers remain relatively stable, the only significant change being in the unfavorable shift on the expensive / inexpensive scale for the A & P. The changers show decided shifts On many items. Noteworthy are the significant shifts in favor— ability toward "fair," "pleasant," "clean," and "friendly" for the Kroger store on Wave II. Kroger, it will be remembered, rmmedo 304 $.30 :03? P b D 1 ¢ 4 A d Fm>ue<>mmlwzoop _ . wtmmmmoommr I) d d u q rmZmzmaxu p p . mamzmsxmzHL I J T q d HZ2.)st A . _ q I 4... :9 zmmzcor q H .JHIZALXJ "P J em .44 U :0... IL . p E 4 52144.52 P P - J .wchfitthD P b 1 d > :2; p P 1 d .2 ..r. ; ‘ 2...? ~.;/ .(k k A— t at” p 1 q I 1.; .1122; n... _ .. x. 1:... w... I «a; x; :1 fit 4;; .45 ...... F r - u ..I Q I . a. I .. 1 Then; IIII...I .473 MK\.(. .H r v i _ /. w I; .... ./ F b P 1 I 4 a . db P d d » mamwm :cozit 1 d . m.,a....7.u.§m..ef. d J . umOJwL d J » c3275??? 4 d d3 so nououo p030m> >08. ad 40ch on» em :uuomddnu: >3 awnmuau .«o nodwounw u- m .m .umh P mmWb<> meZCUp b P mamamoomnmk .1 d 1 1 J P H>~mzmflxm . P p H.2HWZHQXHZHL W 4 1 1 J FZ< Emma QWOJU P h # iP b - J 1 d J P >JQZMHZhZD P P VJQZNHMhb q 4 d J P >H§Q - b P (Mu—JD fi 1 1 d FwZ05 In Joann 05 am :uuouguanoa: an gunman." no noduounm nu o .m .mmh £< C 52,. $.35 :92 p r P P P ‘b ql‘ d d K a d a a . ., . . x / I .. ‘b P *3 «d , “my . P P .., . P HZTWQZCCZQ q‘ d d q d A “2557”; L73; “>507. MAKE 7: p P P » p ‘b 1‘ a d d J M .14. aka..." dmOJU - P p P ‘L 1‘ u d d a - .rJC...,m_.L..m.,_Z P - >1Ha—2dmsz‘p 1‘ q u d ~ . la; > “1%an w.» p P b P P 1‘ d J u d P Huh/:34; T. I p P P a2d3 no nououm U030? >05-d 406% one am : uuomndguLuon: .3 umcflmu mo nomuounm.-- fin .mwh 166 .9. had the largest gain in patronage during the study period. The A & P, which lost the greatest number of customers, shows a very significant shift on the clean / dirty scale in an unfavorable direction, suggesting that the "cleanli- ness" factor may have been a major reason for the patronage loss. Profiles of Customer Groupg, Figs. 5.8 - 5.15 show the profiles of each of the major customer groups toward the major super markets on Waves I and II. Minor stores were excluded due to the small nLg, A brief inspection of this group of charts reveals clearly that each group sees its chosen store as being closest to the "ideal" image, significantly so on those adjectives with high evaluative factor loadings. On those which tended to‘be more denotative than connotative (price and distance) there is still a marked tendency to "see" one's own store as closer and less expensive than the consensus of other groups. For example, the A & P, physically about 50 per cent more distant than the other major stores and so rated by the entire panel, is significantly closer to its customers. The major exception to the "own store-ideal" phenomenon was the universal agreement among the panel that the A & P was the most inexpensive among the stores. It Fig. 5. 8 -- Profiles of rating- by A I: P customers in October. 1960. l .‘ 3 4 L j T / l m L 1 A Y s! I FAIR P I i—‘LLAnAN 1‘ g L fl I CLEAN l l ' . . .. v T f 1‘.‘\1X..\A;l Y [— l \ I , 1* 7 LA .Af'l. P l A I J I _ _ Y \ j l.v}'.)\l'! “. I" 1.. k 1 A J t t T _‘ my; I- II- \ ccmm '; 1'. x. L l L 1 _l r . .v u v ' ,.. —. ‘ j .‘ulutl \._..;.;\.).‘3 [AM LI \ )1 Fig. 5. 9 -- Profile. of ratings by Wrigley customer: in October. 1960. l j #1 1 fl 1% j k 1 I V‘ , j . j r.\1:{ -- L L 1 4 r . ‘ EIL.}~‘- \ 1.} h. . l— L _l ‘ I fl 7-' I.~ I \‘ L 1 #1 r.- . 1 . . j I. 1‘ . :— b‘ J! | L 1 4 r fl CL\J~\1'. _ I K . 'x. g r ‘v ~ - v . - -o In}. .\1 my. .31 . I. ’ L L A 41 pxomausn E. \ .51.. m. .p g ‘ u 4 r211.“ 91.3.55 ‘ 1 Lu‘ u CLASh Km 3 a F’ — LKYJAI. W K1011 "f— K“ '00le — IDLAI S ,1, III — 168 Fig. 5.10 -- Profiles of ratings by A I: P customers in January. 1961. Nll433.“\.)\\i p .JLm \J‘. l(/\L) bAI) db .‘L 1P .JL UNPI-.E.ASAN'I DIRI 'IJYRII-.T.'IJLY .JL Db lAi‘J P \ 1P *‘1P (WI-:3. ‘- «P M 4‘. JL HQSI‘. E L L l J r— r —r fi l'ilc‘f. ‘ I‘ I CCL‘JSER VA '1 IV It. L L J r ‘ Y 7. i (.1911 (Jim‘xofi LOW CL.-\Sb Fig. 5.11 -- Profiles of ratings by Wrigley customers in anusry. 1961. ' L 1 J V f 1 . . .; , ‘ BAD 1 1 v 1 4 V s V ' r"). 1;». V L'L‘JI-‘A u. l k l 4 F __ r T fl NI- LA 3.3.? ‘ UNPLLASA i L L 1 5t . rr.r..«x..‘ 3114. - Yfi 4] 5"‘.;IE..‘.'IZZ.Y ' ..,.‘~.’FRIE.;IL.;LYj : ‘ 4K» . s (‘LuS'rl 1315;.m'r g \'=.. 'IIJL..\'i'L.‘.5l ‘r’rxpm. .SIVE g l ~P D "P «P p b «P {— T v Hum. 55.51%; \‘ consm . A1 vrzg‘ l 1 l 41 T U 'EiImJ cuss Low CLASS KEY .\ a 1’—' xA'HorcAL V‘ “ICL FLY _ KROGER— war-.1- s 1.31. r. — 169 Fig. 5.12 -- Profiles of ratings by Kroger customers in October, 1960. 1 .' 3 l 5 L l l J. J U . 1 V fi fi CHOU \ MAI) L L l J E All-i ‘ L..’E‘AI}< l ' l L l j r f t f 1 j PLLASA." ' warm...“ x": I . . . . -. V . 7 fl C L. L'\ 3‘ ‘ U\, l X L \ —L' 1 j . . v- v ‘v ' ‘.' I 'v‘ 'j EMILAJLY \ UN...“ ..1-\ L L % ‘ % fi - _, ' ' CLUbI'. V 1.x. .A.‘.’i L _I r— ' -’ j lNLXl"E.‘JSI‘.’F A 'r I‘ XE’FIJSIVL L l - 1 L] r V T —_' PROGR rzssx‘; P. \ com LR \ '1 I“. I. l L J . fi V V t of .. . j HIGH CLASD 1.0.5} \ LASS Fig. 5.13 -- Profiles of rstings by National customers in October, 1960. L l l J L] I . 1 (' fill) Rx\l) L l s - ‘ .. . ._ . ' E.\Il\ \ .I?. ‘\L L_‘ ,.. v 1 v - r,LFJ\j' .\ L k}.\l'l-f..‘\.“)f\.vl - I l . fir ' ‘ fl, crlfu‘xfi‘ ‘ n‘- .. \ L L] r V\‘- , . 1 F?I:...TIY.Y .1 . 1 L J I —‘ CLOSE. IL“-.\:.I L L] ‘ .o s - ‘o . I IFJEXPELXBI . 1. EXP}; ...~.1‘. 1; IL _1 PROCIiLSbI‘JE; 1+ . H . }- L - 4 711.421 CLASS Low CLASS KEY A u i’ "" :.'.r.11<~::.»\1. leGLF-Y — KRQCLA — 11" EA 1.. S 'l C-lx I- — Fig. 5.14 -- Profiles of ratings by Kroger customers in January, 1961. l -' 3 4 5 L 1 l 1 LI r v 1 j L(_)L)D ‘ PAL) 1 1 L ‘ V ‘ V 'Vfi 1 1‘ A IR V U“ L‘ I“. I: L L 1 . L j r ~ 1 t T r ' ‘ ‘ ., j k‘LLAb. ‘ UNPLLASAAI ' 1 4 " I j CLLAN ' ..‘U\'1Y L ‘ L J r , . ,, ' f . .. j 1‘ 111145! L)14Y UP‘1' 1111‘..~L.‘ 141 L ' CLUSL‘ p I 1..'1;.X1’E.‘JSI‘.’ 1. 'F’ I‘RCCR ESSIV E L L HIGH CLAb‘b 1. JL 12.1.8 1A Bl L L. L T fi CONS ER VA 1 I'.’ I: J fl LOW C l .--\ 55 Fig. 5.15 -- Profiles of ratings by National customers in January, 1961. L 1 1 _ 1 h. .‘ t Y ' . ‘ 1 (.1.- 1. I, 135.1) L L 1 1 J \‘ ' f \ j —v.' .. 1 'E.\IR _\ “Ian. ' 1 1 L ' 1 L1 ‘ V V V - . j PLEA ~ I “ UTJ’LPASA... L 1 _1 C1,}... ' ' V, :rn: .yfi L 1 1 1 L] r'" , r I , . .. l }:l.I. . ‘. [Mfg-.11”, 1 1 1 A _l miOCRESbM k. ‘51-... H. . L L | Ll V j 7mm CLASS .'\ :1 1' W R101. 1. Y — 13 14A 1- S1L1\1 L.’:\'1 103.31. KRUGLR 171 will be remembered that the experts found A & P prices to be virtually indistinguishable from those of the other super markets. The profiles are a testimonial to the A & P's successful projection of a "price image" as described in Chapter IV. For the clearly evaluative adjectival pairs, in every case the means of the store groups favored the chosen store. Customers tend to think of their preferred store as being the most "fair," the most "good," etc., in harmony with valued norms. In Fig. 5.1, where the total sample mean for Wrigley on the clean / dirty scale shows an extreme shift toward "dirty,” customers of that store rated it as the cleanest. In almost all cases the null hypothesis that the four major store groups were random samples of the same population was rejected. Variances were smaller for customers rating their own own store, suggesting a greater consistency of attitudes. The smaller variance is also due in part to the fact that such intense scores are limited by scale extremes in the measuring instrument. The profiles reveal each store group to have general— ly stable attitudinal patterns. There are, however, distin- guishable shifts which can be seen by comparing the profiles of each store for the two waves. The A & P tended to lose the t f g u 172 favorable position it held relative to the customers of the other stores, being replaced by the Kroger store. This is clearly evident in the National profiles. On the other hand, Wrigley customers rated the A & P about the same on both waves, but show significant changes in a favorable direction toward Kroger and National. The stores with the most favorable profiles after "own" store suggest the direction in which changers may shift. Only a few of the inferences which are suggested by profile analysis have been discussed. It would seem that the semantic differential data is a useful method for quantifying the abstract "image" held by people toward economic objects such as super markets. There are, of course, numerous modifications that could be introduced into the instrument. Mindak (1961, pp. 28-33) suggests the use of phrases and nouns in place of the adjectives on the scales, the construction of scales "tailor-made" for specific prob- lems through content analysis of the advertiser's own and competitive advertising, and word association tests with consumers . Three-dimensional Analysis As an example of still another method for evalu— ating semantic differential data, Figs. 5.16 — 5.23, using e! '1‘. A a P . NATIONAL WRIGLEY . pnmcg KROGER SCHMIDT . IDEAL STORE . DISTANCE SCALE Fig. 5.18 -- Three—dimensional image by Kroger customers, October, 1960. Q l'r' Q Q Q) . 2.... 3" O O Q $4 4 a e s .1 6 5 4 3 2 I EXPENSIVE INEXPENSIVE Fig. 5.19 -— Three-dimensional image by National customers, October, Q 1T 1960. Q Q K3 2“ . 31r- Q ‘1 ‘1‘ 4 .2 L s 4 6 5 3 l EXPENSIVE INEXPENSIVE 177 three of the nine scales, show relationships in three— dimensional form. The three adjectival scales were selected as being the most logically representative of the nine adjectival pairs used in determining customer attitudes. While we have already seen that there is considerable inter-relation- ship among scales, it was assumed that price and location might be indicators of more "rational" denotatively oriented dimensions, while "good-bad" could represent the connotative or evaluative dimension. A three-dimensional image is shown for each store group for Waves I & II. The dimensions can be readily noted by inspection of the figures. The higher and further to the right a store-concept appears the more favorably it is viewed for the respective scales. The "close—distant" scale is indicated by the size of the circle representing a concept. The larger the circle, the "closer" the concept is rated by the group. Again, it is apparent that each store group sees its own store as most nearly resembling the "ideal" store. More importantly, the figures show the alternative store most likely to benefit from any changes in patronage or to be favored for secondary shOpping. For example, the A & P customers reveal no pronounced preferences for other stores 178 in October. National and Schmidt's have a slight price ad- vantage, although Schmidt's has distance disadvantage. Kroger is seen as most "good" and as closer. Inspection of the Jan— uary "changers" from A & P show that five went to National, three to Kroger, two to Schmidt's, and two to Wrigley. The Wrigley group (Fig. 5.ID favors Kroger and A & P, but the A & P has a disadvantage in distance. The January "changers” from Wrigley revealed five of seven transferring to the Kroger store. Kroger customers indicate that National and Wrigley were about equal on all dimensions in October. The six "chang- ers" split between these stores. National customers preferred the A & P in October, although Kroger and Wrigley are seen as closer. The "changers" split two to the A & P, two to Wrigley, three to Kroger, and two to minor stores. The charts for Wave II show small but distinct shifts in the attitudinal structures of the store groups. Aggregate tendencies have changed slightly from October. The A & P and National customers now clearly prefer Kroger as second choice. Kroger customers lean to National, while Wrigley customers tend to favor the A & P. Overall, perhaps the most noteworthy aspect of the structures is their basic stability. While the attitudes of some individuals changed considerably within the short period of time, attitudes of the panel were quite consistent, 179 although revealing dynamic tendencies in the slowly shift- ing patterns. The fact that store attributes changed very little during the study period, according to the experts, suggeSts that external communications were a powerful factor in accounting for the shifts that took place. For example, the three-dimensional images for October would suggest that the A & P would benefit the most from changes in patronage despite its location dis- advantage. As will be noted in the next section, actual patronage shifts did not bear out the assumption that this strong price belief would favor the A & P. One reason will be seen in the findings from the linkage analysis. A second factor was a shift in attitudes toward the A & P, possibly as a result of communication pressures, such as the use of advertising by the Kroger store in the State News following Wave I. As can be seen in Figs. 5.22 and 5.23, the customers of National and Kroger no longer viewed the A & P as favor- ably in January as they did in October. Especially notice- able is the significant shift toward "expensive" on the inexpensive / expensive scale. LINKAGE ANALYSIS OF THE SUPER MARKET IMAGES The purpose of the next section of this chapter is to investigate a technique for clustering the multitudinous 180 variables the researcher inevitably encounters when study— ing consumer behavior into a comparatively small number of representative variables. It was reasonable to conjecture that the number of underlying images (profiles of stores) might be fewer in number than our measurements would indicate. According- ly, the profiles for each of the seven concepts obtained from the four major customer groups were correlated with each other. The 28 x 28 matrix of intercorrelations was computed after first collapsing the adjectival scales. The method used to determine how the various profiles clustered together was McQuitty's "linkage anal- ysis," described in Table 3.1, Chapter III. Figure 5.24 reveals the three clusters of "families" found in the pilot study. Type I shows the customers of the major super markets tended to look upon the minor super markets in much the same way. This phenomenon could be labeled the "other store" image, in that the two stores involved were characterized by very low patronage from the study group. Type II shows that all customer groups were in high agree— ment on what constitutes an "ideal" store; closely linked to the ideal image were their own chosen store images. Still attached to this cluster was a secondary cluster consisting of the images of the A & P held by the customers of the other 181 Fig. 5. 24--Clusters derived by linkage analysis of the images held by the customers of the ”Big 4, " January, 1960, pilot study. _1_ ~ .97 ”37% W/P-=>K/P —->N/P -—->A/Si———=K/S W/sa N/S .50 .69 .74 .99 .98 \fléH—t—a—Jj II W/A—9 A/A éA/ISK/I 2:2W/14—W/W .87 .97 1~ .98 f\98 .96 .63 .93 .96 | l I A/K K/K N/I “N. .86 .93\ K/A<——-' N/A A/P—-?N/K<——"—’N/N .95 .53 .95 f. 80 W/K:_—_>w/N<———N/w w;—_"."K/w .91 f\. 86 .85 82 .75 . I I K/N A/N £215 Key: First letter indicates customer group making the rating; second letter indicates the concept being rated. For example, A/A indicates A 8: P customers rating the A 8: P. A - A 8: P* -——" ”brother/sister" relationship W - Wrigley"I "fir st cousins" K - Kroger* ———-> "first cousins” N - National * P - Prince (AA—4+.) "closest friends" S - Schmidt 1 .. Ideal Super Market at: ”Big 4" supermarkets 182 three major stores and the image of Kroger held by the A & P customers. This cluster could be labeled as the "preferred store“ image. Type III indicates that the groups saw Wrigley and National similarly, but in a less favorable light. This cluster was labeled the "problem store" image. Fig. 5.25 reveals the clusters found by analyzing the data obtained on Wave I. The three generic types appeared as in the pilot study; however, there appears to be a reshuffling of the super markets within the categories. The first type de- rived has become a "shopping center" cluster, indicating the panel looked with greater favorability upon the three super markets located in the shopping center. The panel still agrees on an "ideal" image, although the shopping center stores are no longer linked to the group. On the basis of the typal relevancies, it was predicted that Kroger and Wrigley would be the main bene- ficiaries of any changes that took place. The findings from Wave II generally supported the hunch, although Kroger gained a greater share than anticipated from the typal structure: Changers to A & P Kroger Wrigley National Other From A&P - 3 2 5 4 Kroger 3 - 3 none none Wrigley l 5 — 1 none National ;_ 3_ ;_ ;_ y; 6 ll 7 6 6 183 Fig. 5. 25-—Types derived from matrix of intercorrelations between profiles of each store by customer groups of the major super markets, Wave 1, October, 1960. Type I - (Shopping Center Cluster) .98 .99 .97 .98 K/W —\ W/Kv—éN/W N/N-A w/WF-‘K/K .83 .9 .92 N/K .96 .97 .98 A/W x———"—A/K A/N —"s—-K/N '91; .95 .90 .95 .95 '94 A/S V—‘K/S H W/S N/SV—W/N Type 2 (Prince Cluster) .97 .99 .95 A/P.__>. W/PgK/PV— N/P Type 3 (Ideal Cluster) .98 .96 .94 .94. .93 MI =K/I ‘___w/1‘__ A/I V..._\__s._.._>A/A =W/A¢_K/A 1... N/A KEY: The first letter indicates customer group making the rating. The second letter indicates the concept being rated,- e. g. , the symbol "A/A" indicates the rating of the A 8: P by A 8.: P customers. A - A 8.: P P - Prince ——-“__ "brother/sister" relationship W - Wrigley S - Schmidt K ' Kroger I - "Ideal" "first cousin" " N - National **"closest friend" " 184 The Schmidt store has left the "other" category and now appears to be a more suitable alternative choice. In other words, this super market has achieved a more comparable degree of favorability with the "shopping center" stores. By looking at the profiles, Figs. 5.1 to 5.3, it can be seen that Schmidt's had overcome an unfavorable price image, although the distance factor was still a barrier. The earlier image had apparently been due to a belief that the ultra-modern building housing this store meant higher prices (Schmidt's was relatively new at the time of the first study). An attempt was made to find out how serious the distance factor was in preventing more Villagers from patroniz— ing Schmidt's following Wave 1. One revelation was the fact that many respondents who had visited the store and had favorable attitudes had taken a route that was much further than the route taken by regular patrons. In short, if this super market would provide a simple map in their advertisements showing "short” routes, it is conceivable they could increase their share of the market among the Villagers. The strong A & P image that was found in the pilot study clusters shows signs of weakening in the October (Wave I) clusters. It suggested that the A & P was losing its favored status, giving a hint of the patronage loss that was to occur between October and January. 185 SUMMARY The findings reveal the semantic differential scales used in the study to have a unique ability to isolate factors important in store selection. The instrument can provide graphic images, showing both direction and intensity of feelings toward competitive stores. The profiles also serve as useful indicators for more intensive investigation in problem-areas. They seem particularly useful in comparing consumer preferences among classes of consumer products, as well as for products within the classes (as in this study). The profile data can also be used in different forms, such as the three dimensional graphs, which can further define factors. It is assumed, of course, that such investigations will take place over time to provide maximum information for management decision—making. For larger studies, the use of linkage analysis appears to be a very feasible procedure for developing scales and concepts for specific problems. It also provides a convenient method for extracting clusters showing underlying images. The principal advantage of linkage analysis as a method is its ability to cluster variables into types rapidly and objectively, whether consumers, products, or items. The differential used included only nine pairs of adjectives. No doubt there are many others relevant to the images 186 of stores. There appears no reason which would prevent the use of symbols other than adjectives to derive additional meaning. Yet, as used in this study, the instrument as con- structed provided much valuable information for the problem investigated. Researchers in applied areas could eventually catalogue customer and product types. This could be a very valuable aid in explaining and predicting a number of market- ing problems. Studies using the methodology described would be inexpensive and readily adaptable to local as well as national studies. The findings revealed that the attitudinal patterns of the panel were quite stable, although discernible trends could be noted. The most striking feature of the profiles is the fact that store customer groups clearly "see" their own store as having qualities which the "ideal" stereotype incorporates. In other words, when choosing among relative- ly identical products, the housewife may make a choice as a result of predispositions and verbal appeals from face— to-face and mass media sources; then, if her experience is congruent with her expectations, she attaches the set of desirable attributes to her choice. When she makes a change to a competitive product, the entire cluster of attitudes shifts. 187 The verbal appeals must correspond to her experience, however, as can be seen in the Wrigley profile changes on the clean / dirty scale. The respondents reacted swiftly to differences in attributes when clearly recognizable. The concept of "image" as developed in this chapter, and the methodology for defining it, appears to generate‘ valuable data for guiding management in taking more effective reinforcing or remedial action, such as a shift in advertis— ing strategy. 188 CHAPTER VI TWO-STAGE ATTITUDE CHANGE The intent of this portion of the investigation was to assess the effects, if any, of mass and face-to-face com— munication concerning a socio-economic object, such as a super market, upon consumer attitudes and behavior. The first section of the chapter sets forth the theo- retical formulation underlying the concept of "two-stage atti- tude change" which hypothesizes that the position of a consumer on an attitude continuum as well as actual patronage behavior will change as a function of the combination of messages and predisposition. The second section explicates the methodology used. The findings are discussed in the last section of the chapter. THEORETICAL FORMULATION From the paradigm of the buying act described in Chapter II, buying behavior is seen as in expression of spec- ific attitudes reflecting the varied influences of the consum- er's personality and the situation in which she has her being. Singled out as primary influences in the situational field were the consumer's group, the communications received through the mass media and face-to-face channels, and attributes of the super markets. 189 One of the problems encountered, that there is not a one—to-one correlation between consumer beliefs and objective facts about a super market, was described in the last chapter. Despite her every-day familiarity with super markets aVailable to her, there may be a wide discrepancy between her beliefs and the objective facts about the stores. This factor will not be considered separately in the present analysis; instead it is lumped into the more general concept of predisposition. The primary focus will be upon the influence of the two primary communication channels, face-to-face and mass media, as screened through the "self" messages emanating from the individual's predispositions. The difficult problem of designing this portion of the study was aided by the excellent pathfinding done by Katz and Lazarsfeld (1955) who suggest the predisposition-group variables are more effective than the mass media, and that of Deutschmann and Pinner (1960), who found that political influ- ence is dependent upon mass media for arousal, and conclude that attitude change may result from personal influence, but only when it is added to the messages first received through the mass media which initiates the change process. The basic approach followed that suggested by Lazars- feld, §§_§1, (1948, p. 151) in their "two—step flow of communi- cation" hypothesis; therefore, indices were needed for: 190 1. Exposure by respondents to advertising messages in the local media. 2. Exposure by respondents to face-to-face communi- cation about any of the super markets, and whether favorable or unfavorable. 3. Predispositions of the respondents toward the available super markets, in the form of attitudes as measured by the semantic differential instru- ment and in the form of stated store preference and patronage. Deutschmann and Pinner (1960) point out two reasons which might explain the anomaly resulting from the fact that most field studies have not supported laboratory experiments which show mass media messages can change attitudes: (1) lack of sensitivity in the instruments used for measuring attitudinal changes in the field, and (2) memory loss occurring due to the delay between receipt of the messages and the measurement of effect. The present design thus included a sensitive measure of store attitudes in order to account for the small, but real, changes that may take place, in a natural situation. The second factor, memory loss due to measurement delay, was partially met by conducting the field operations on Wave I within a short time after the new residents in the sample were first exposed to advertising and had just begun to investigate super markets in the area. A modified version of Deutschmann's (1960) model of the two-stage attitude-change process was used, involving the following assumptions: 191 1. Any message will have an effect in its intended direction. 2. Any face-to-face message will be as powerful as any message transmitted by the mass media. 3. Any predisposition will operate in a manner simi- lar to a message; in effect, this implies the in- dividual "sends himself" a message, with the same power as a face-to-face message. 4. The combination of predispositions, mass, and face-to-face messages are additive. While these assumptions enable the researcher to specify the probable effects of two-stage flow communications situations, given a set of messages relating to some social Object, the focus herein specifically is upon super markets. A diagram utilizing arbitrary weights illustrates the model: +1 (.Mags message_‘ 0 +1 Lieceiver predisposition |=weight=2 [Face-to-face message] It was hypothesized that the position of the consumer- receiver on an attitude continuum concerning a given super market will move as a function of the combination of messages and pre- dispositions. Take, for example, a new resident who patronized an A & P supermarket prior to her arrival in Spartan Village (predisposition is plus). She reads an advertisement in the newspaper for Kroger, and talks to a neighbor who strongly recom- mends Kroger (messages are both minus). 192 -l| advertiggment 5 +2 Lreceiver predispositionyj=weight=—1 —2 face—to-face , recommendation Her attitude toward A & P will shift to a less favor- able position. This action schema, while elementary, has the merit of being operational with reference to the factors under inves- tigation. In particular, the complex of variables associated with the concept of predispositions can be accommodated. The event, store selection, is one that can be considered crucial to the respondent, yet relatively uncontaminated by the process of rationalization where the respondent says one thing and behaves differently - the distinction between "real" and "ideal" behavior that plagues the survey response. The next section will describe the construction of the several indices used in analyzing the observations relevant to communications effect. Insofar as possible, the requirements for making an index of high quality were observed: (1) the vari- ables measured are clearly defined, (2) the data is relevant to the issue, and (3) the index formula itself was thoroughly tested and revised so as to minimize self—induced distortions in analysis of the data. 193 METHODOLOGY In order to test the entire complex of hypotheses surrounding the attitude-change process, six indices were computed. Five of the indices were measures of message pres- sures, the sixth was a measure of the dependent variable, attitude change. Correlation methods were used for testing. 1. Index of Predisposition "A" - Two distinct measures of predisposition toward chosen super markets were used. Predisposition Index ”A" was an objective measure, obtained in wave I by arbitrarily weighting answers to ques- tions concerning: (1) super markets patronized by respondent prior to moving into Spartan Village: (2) the super markets patronized by respondent "last week;" and (3) those super markets rated by the respondent in order of preference in the East Lansing Shopping area. The weighting scheme is shown below: Prior patronage of same chain 10 Ranked as first choice 8 Ranked as second choice 6 Ranked as third choice 4 Visited but not ranked 2 Shopped last week, major purchase 4 Shopped last week, minor purchase 2 Maximum score possible 22 (any one store) Minimum score possible 0 " H n 194 These weights, while arbitrary, were derived on the basis of earlier findings in the January, 1960, pilot study. They have face validity to the extent that the distinctions they provide correspond with judgments made by other observers rating the importance of the various components of consumer loyalty, as well as the "common sense" evaluation made during the earlier study. As was shown in Table 4.4, Chapter IV, some 60 per cent of the panel preferred the same chain locally as pre- viously patronized in their former place of residence. The other two categories, respondent's rating of stores and "shopped last week," were assigned weights after a careful analysis of several factors included in the October questionnaire. These were designed to evaluate the several influences surrounding store choice. (See Appendix C, Wave I Questionnaire, 1.1-7.1 and corresponding descriptive data in Chapter IV). For example, two-thirds of the decision- makers shopped "last week" only at the superrnarket rated as preferred. Overall, 90 per cent either purchased all or most of their groceries at their first choice store, and about 10 per cent made their purchases at a store other than first choice store. While some women will rate a store as "first choice," they often indulge in "looking around" or find it expedient to do shopping in another store. About one-third 195 of the panel regularly shopped in two or more stores. Visiting any store was considered comparable to receiving a favorable message, while visiting plus a ranking, produced a higher pre- disposition score. To obtain a measure of the relative strength of the consumer's predisposition toward her preferred store, a pre— disposition score was constructed for each major super market. Only customers of the four major super markets were evaluated, reducing the effective g_for this portion of the study to 145 respondents. Of the original 189 persons on the panel, twenty moved away, while the remaining fourteen preferred a non-major store on one or both waves. Predisposition "A” was obtained by using as the base the super market which each customer stated was her preferred store in October. The predisposition score for each of the other three stores was subtracted from the preferred store score and the remainders summed. The total became Predis- position "A." As an example, the computations for a respon— dent who preferred the Kroger super market in October are shown below: 196 Predispgsition Weights A & P Wrigley Kroger National Prior patronage O O O O Ranking 6 O 8 4 Shopped "last week" 0 O 4 O 6 O .1; 4 Difference 12-6 = 6 12-0 = 12 12—4 = 8 Predisposition "A" = 6 + 12 + 8 = _g The resulting weights ranged from a minimum of zero to a maximum of sixty-six. Only one respondent had a score as low as zero. Three respondents attained the maximum score, indicating they had previously patronized the store, selected it as first choice, did all their shopping there "last week," and had neither ranked nor visited any other store since moving to the Village. TABLE 6. l PERCENTAGE OF SUBJECTS ACCORDING TO OCTOBER PREDISPOSITION "A" TOWARD THE FOUR MAJOR SUPER MARKETS (n=l45) Index of Predisposition (weak/Low to Strong/High) Per Cent 0 ‘ 9 5.5 ( 8) 10 - 19 13.1 (19) 20>- 29 34.5 (50) 3o - 39 16.6 (24) 4o - 49 2.1 ( 3) 50 - 59 16.6 (24) 6O - 66 11.7 (17) 100.0 XEEEEQmers of other stores: 24) 197 2. Index of Predigposition "B" — The second measure of predisposition was based on respondent attitudes, operation— alized as the scores received by each store concept on the semantic differential instrument. The index was computed as follows: a. The mean score for each respondent's preferred store was obtained by summing across the nine scales and dividing by nine. Each scale was weighted from one to seven, with the most favor- able score being one, the least favorable, seven. b. The scores received across all scales for each of the three other major stores was summed and divided by 27. c. The mean for the preferred store was subtracted from the standardized mean for the other stores. The score thus obtained was termed the "attitude difference score." d. The "preference" scores were arrayed and assigned an arbitrary index number, as shown in Table 5.2. For example, one respondent received the following SCOI‘BS: A & P mean score 2.1 (Mean - vi mhx wri le " " 2.8 g y Where 2_:: difference score Kroger " " 2.2 :: own store $8 National " " 4.0 zlzz other stores Then: D =‘Xb - - 2.8 + 2.2 + 4.0 =2.l- 3 =2.1-3,1 = -.9 198 By definition, a favorable preference score would have a minus sign. These were reversed in the final indices. It can be noted in Table 6.2 that the difference scores are normally distributed. Support for the validity of these preference scores is provided by the fact that g; per cent of subjects had a differential favorable to their stated preference. The mean preference was 1.1 scale points. TABLE 6.2 PERCENTAGE OF SUBJECTS ACCORDING TO OCTOBER PREDISPOSITION "B" (ATTITUDE PREFERENCE SCORES FOR CHOSEN STORE VERSUS OTHERS) (n=l45) Attitude Preference Score Per Cent -l.00 to .00 7.6 (11) 0.01 to .50 18.6 (27) 0.51 to 1.00 17.9 (26) 1.01 to 1.50 27.6 (40) 1.51 to 2.00 20.0 (29) 2.01 to 2.50 6.2 ( 9) 2.51 to 3.00 2.1 ( 3) 1.: O O O 3. Indexeof Face-to-Face Pressure - Questions 4.1 and 4.2 (see Appendix, Wave II QUestionnaire) were included in the reinterview questionnaire to assess the impact of face-to—face influence toward the "Big Four" stores. The lesser supermarkets were not included due to the small number of customers. It will be noted that the 199 questions incorporated two time periods, last week and past three months. For each store the respondent could show direction of conversations as favorable, mixed, or unfavorable. Arbitrary weights were then assigned each store, as follows: Direction of Conversation Weight Talk favorable last week plus 4 Talk favorableypast three months Favorable/Mixed - either time plus 2 FavorabngNo Talk — either time Unfavorable/Mixed - either time minus 2 Unfavorable/No Talk - either time Unfavorable/Unfavorable — both times minus 4 All other combinations 0 To obtain a measure of relative pressure from face— to-face conversations between the stores, it was necessary to construct a method of evaluating all four "Big Four" stores. This was done by using the super market stated by each respon— dent as her preferred choice in October as a base, as follows: . Kroger . A & P W N t rigley (Preferred) a ional Face—to-face _ score -2 2 2 0 Difference 2—(-2)=4 2-2=O 2-O=2 Total = plus 6 200 This enabled both favorable and unfavorable differ- ences to be taken into consideration. The value of plus 6 indicates face-to-face presSure in favor of the preferred store. For the total sample, the scores were as follows: TABLE 6.3 PERCENTAGE OF SUBJECTS ACCORDING TO FACE-TO—FACE PRESSURE SCORES TOWARD OCTOBER PREFERENCE STORE ON JANUARY REINTERVIEW (n=l45) Pressure weight* Per Cent Plus 10 and over 15.9 (23) Plus 6 to plus 8 19.3 (28) Plus 2 to plus 4 15.2 (22) No pressure (or balanced) 26.9 (39) Minus 2 to minus 4 15.9 (23) Minus 6 to minus 8 4.8 ( 7) Minus 10 and less 2.1 ( 3) 100.0 *Plus scores indicate pressure favorable toward preferred store (using October preference), minus scores indicate pres— sure received against preferred store. Of the total sample of 169 on second wave, eighteen respondents were not cus— tomers of major super markets and are not shown. Briefly summarizing the findings from the face-to— face index, the following is noteworthy: (l) the majority of respondents received messages supporting their preferred store. (2) only about one in five engaged in conversations which were against the preferred store. 201 While these findings are in agreement with small group research findings indicating that conversation occurs among like-minded persons (Festinger, g£_g1,, 1950; Newcomb, 1943, 1953), the low percentage in (2) suggests selective recall on the part of the respondents (Krech and Crutchfield, p. 107). 4. Index of Advertisigg Pressure — Each of the "Big Four"super markets was analyzed separately by each of the four major media (newspaper, radio, TV, mail) to which the respondents had been exposed. As with the face-to-face questions, the respondents were asked to recall any exposure to each of the four types of advertising "last week," and "past three months." (See questions 5.1 thru 8.2 of the Wave II questionnaire.) The questions were scored as follows for each of the "Big Four!" Newgpapegg: Last week - Saw no ads 0 Read less than half 1 Read more than half 2 Past Months — Saw no ads 0 Saw once a month Read less than half 1 Read more than half 2 Saw once a week Read less than half Read more than half 3 N Range of possible scores: 0 — 5 202 Radio, Televigion, and Mail Last week (heard or read) - Yes 1 No 0 Past three months Yes 1 No 0 Total possible score, any one medium 2 Grand Total possible,yper store .11 It will be noted that newspaper was accorded a higher weight than the other three media. This was to reflect its much greater usage by these supermarkets providing opportunity for more exposure. No assertion is made that newspaper adver- tising is more effective. An example of the effectiveness of a particular medium was provided during wave I. Two of the major super markets had sent out mail pieces the week previous. Patronage figures showed both stores had an approximate ten per cent increase in business, drawn from customers from the stores not sending mail advertising although all stores had used the same amount of newspaper advertising. Following the scoring of each medium, it was neces- sary to total across the media to derive a total media score for each respondent for each store. The range of scores could thus vary from O - 11 for any store. The advertising pressure score was then computed. The objective was to obtain some index of relative pressure 203 in the direction of chosen store that the advertising might be exerting upon a respondent. Since all advertising messages were, by definition, "favorable" each ad was given a plus value. But since the respondent might be exposed to the advertising of several stores, the weighting scheme was designed to express the differences in exposure as is shown in this example: For a respondent who was a Kroger customer in October: A & P Wrigley Kroger National Difference Newspaper score 2 O 4 2 Difference (4-2) (4-0) (4-2) plus 8 Radio 0 0 l 0 Difference (1-0) (1—0) (1—0) plus 3 TV 0 2 0 0 Difference (0-0) (0—2) (0—0) minus 2 Mail ' 2 0 1 0 Difference (l—2) (l-O) (l-O) plus 1 Total Difference: plus 13 The distribution of advertising weights is shown in Table 6.4. It will be noted that two-thirds of the panel re- ceived an advertising pressure score favorable to their "own” store. Only one of five was found to have a net "other" store pressure. The finding is in general agreement with studies on selective perception (for review of studies, see Klapper, 1957), and closely parallels the finding for face-to-face recall. 204 TABLE 6.4 PERCENTAGE OF SUBJECTS ACCORDING TO ADVERTISING PRESSURE SCORES TOWARD PREFERRED STORE IN OCTOBER WHEN REINTERVIEWED IN JANUARY (n-l45) Pressure weight* Per Cent Plus 16 and over 4.8 ( 7) Plus 11 to plus 15 13.1 (19) Plus 6 to plus 10 17.9 (26) Plus 1 to plus 5 31.7 (46) No pressure (or balanced) 12.4 (18) Minus 1 to minus 5 15.2 (22) Minus 6 to minus 10 3.3 ( 5) Minus 11 and less __l;§_ ( 2) 100.0 * Plus scores indicate advertising pressure favorable toward store preferred in Oct., minus scores indicate unfavorable. 5. Index of Gross Change Pressure - To obtain an overall index of communication pressures the arbitrary weights computed for each respondent for Predisposition "A," Face-to— Face, and Advertising were summed and formed into a frequency distribution containing six class intervals. The groupings were found to approximate a normal distribution, as shown in Table 6.5. The six classes were assigned index numbers from 1 (most favorable to preferred store in Wave I) to 7 (most un- favorable to preferred store in Wave I). 205 TABLE 6.5 PERCENTAGE OF SUBJECTS ACCORDING TO GROSS CHANGE WEIGHTS TOWARD PREFERRED STORE IN OCTOBER WHEN REINTERVIEWED IN JANUARY (n=l45) Gross Change Score (favorable to unfavorable toward October Per Cent preferred store) (Favorable) 3 - 7 11.2 (16) 8 - 9 15.4 (22) 10 - 11 22.3 (32) 12 - 13 25.9 (38) 14 - 15 16.8 (25) (Unfavorable) l6 - 21 8.4 (12) 100.0 6. Attitude Shift Index - The attitude preference scores (Predisposition "B") computed from the semantic dif— ferential scores in October were used as the base for com- puting the Attitude Shift Index. A store preference score was computed for each individual from the January data and subtracted from the base preference score. The computation of the January preference score was identical with that out- lined for Predisposition "B." The difference between the October and January preference scores showed the changes in attitudes occuring during the three—month interim. For example: Subject No. 23 October preference score = 2.63 January preference score = -.67 Net Change —3.30 206 Table 6.6 shows the distribution of the January preference scores. When compared with Table 6.2, it can be seen that a definite shift in the direction of less favorable attitudes toward the preferred store had taken place between October and January. Table 6.7 shows the distribution of the net differences in attitude preference scores between the two waves, the "Attitude Shift" index. TABLE 6.6 JANUARY, 1961, ATTITUDE PREFERENCE SCORE: OWN VERSUS THE OTHER MAJOR SUPER MARKETS (n=l45) Score (low to high toward store preferred in October) Per Cent -1.00 to 0.00 14.5 (21) .01 to .50 17.9 (26) .51 to 1.00 26.9 (39) 1.01 to 1.50 19.3 (28) 1.51 to 2.00 17.2 (25) 2.01 to 2.50 3.4 ( 5) 2.51 to 3.00 0.7 ( 1) 100.0 For the sample as a whole, the mean shift in the attitude preference scores was -.267. The standard error of the mean was .056, showing the downward shift to be highly significant. This was at first thought to be a function of the changers in the sample who were now patronizing a dif- ferent store, thus an unfavorable shift toward their October 207 TABLE 6.7 DIFFERENCE BETWEEN OCTOBER, 1960 AND JANUARY, 1961 ATTITUDE PREFERENCE SCORES FOR OWN VERSUS THE OTHER MAJOR SUPER MARKETS (n=l45) Difference Score (low to high toward store Per Cent preferred in October) -l.50 and lower 4.1 ( 6) -1.00 to -1.49 12.4 (18) - .50 to - .99 17.2 (25) 0.00 to - .49 36.5 (53) .01 to .50 17.9 (26) .51 to 1.00 8.3 (12) 1.01 to 4.00 3.4 ( 5) 100.0 Mean shift = -.27 Standard deviation = .674 UK = .056 preference could be expected. The mean shift downward in the changer's scores on the Attitude Shift Index proved to be -.49, highly significant. When the non-changers were assessed separately, the mean shift was -.20 of a scale, also beyond the .05 level of significance. Between changers and non- changers, however, the shift downward for the changers proved to be significantly greater than could be expected by chance. The shift downward for the non—changers was not un- expected inasmuch as the attitude shift computations did not take into account that respondents already at a maximum posi— tion relative to the store preferred in October had no space 208 into which to move following favorable communications. This "ceiling effect" has been noted by Deutschmann and Pinner (1960) in their study of voting intentions. It is conceivable, therefore, that if those who had maintained their positions of extreme favorability (a most desirable effect from the standpoint of the advertiser) had been given an added weight, the mean shift downward for the non-changers would have been smaller. To ascertain if the shift was due to learning on the part of the new residents, a chi-square was computed for new and old residents. No difference was found. This suggested that it might be a function of the stores them- selves. A chi-square for the four major store groupings was computed and found not significant, with all groups showing less favorable means. However, when the A & P group was tested against the other three, a significant difference ‘was found. This suggested that the customers of A & P were less satisfied with their choice overall. Summary of Indices - In order to provide a brief review of the indices and their purposes, they are summarized as follows: A. Independent Variables. l. Predisposition "A" - provides an objective rating-type measure of each respondent's at- titudes toward the super markets in the area. 209 Predisp03iti'on "A" is used to obtain a measure of the influence of "self-messages," and is one of the three message components in the "Gross Change Weight" formula for evaluating total communications influence. Face-to—face Pressure - provides a measure of the influence of interpersonal messages received by each respondent relative to the super markets. Advertising Pressg£g_- provides a measure of the influence of the advertising messages projected by the super markets in the various media during the interim period between waves. Gross.Chanqe Weight - provides a summation of the three message components above and is used to observe their influence on both attitudes and actual patronage. Dependent Variables. 1. Attitude Shi§p_- provides a measure of the change in attitudes held by each respondent and was obtained by subtracting the differential attitude between the preferred store and the other super markets in Wave II from the base differential attitude (Predisposition "B") found on Wave I, as computed from the semantic differential data. Actual Patronage Behavig£_- provides an all—or— nothing index of behavior by ascertaining the actual changes in super market patronage which occurred during the interim between the two waves and constitutes an ”acid test" of the predictive powers of the indices computed. Combined Variable — Predisposition "B" provides the base differential attitude for "attitude pref— erence" score, for computing the "Attitude Shift," thus is used as an independent variable. It is a dependent variable in the sense that it could also be used on the first wave to measure the attitudinal effect of the message pressure com- puted from the first wave, although not intended to show the dynamics of change. 210 INTERACTION OF VARIABLES - FINDINGS A chi-square analysis of the panel across pertinent classifications and indices was executed. Wherever possible, the variables were dichotomized. Eleven primary variables were cross tabulated against each other and various other selected classifications and indices. Table 6.13 summarizes the pattern of differences (chi—square significant beyond the .05 level). The numerical symbols below are used in the table. Sympol Primary Variables 0 Changers v. non-changers 1 New residents v. old residents 2 Life-cycle - by four categories 3 Socio-economic - by six categories 4 Newspaper readers v. non-newspaper readers 5 Opinion leaders v. non-opinion leaders 6 Gregarious respondents v. non-gregarious 7 Predisposition "A": high scores v. low scores 8 Predisposition "B": high scores v. low scores 9 Face-to-face pressure: high scores v. low scores 10 Advertising pressure: high scores v. low scores 11 January attitude difference: high scores v. low scores 12 Attitude shift: high scores v. low scores 13 Gross change weight: high scores v. low scores Other Variables 14 Four major super market customer groups 15 A & P v. Kroger, Wrigley, and National customers 16 Kroger v. A & P, Wrigley, and National customers 17 Wrigley v. A & P, Kroger, and National customers 18 National v. A & P, Kroger, and Wrigley customers 19 Store selection factors: lst choice attributes 20 Ranking of "self," face-to-face, and advertising according to stated importance as sources of information about super markets 21 Ranking of the four mass media according to stated importance as sources of information about stores 211 22 Ranking of face-to-face influences according to stated importance as sources of information about stores. 23 Shopped at one store "last week" v. more than one store. The differences are discussed below in the same order in which they appear in Table 6.13: 0. Changers v. non-changers. The changers in the panel differed significantly from non—changers as follows: a. b. Changers tended to be opinion leaders. Changers had lower predispositions ("A") toward super markets preferred in October. Changers had lower attitudinal difference scores toward preferred store in October, as computed from semantic differential given in October (Predisposition "B"). Changers showed greater exposure to face—to-face influence unfavorable to store preferred in October. Changers showed greater exposure to advertising for stores other than store preferred in October. Changers had larger mean difference scores in unfavorable direction for preferred store in October on semantic differential given in January. Changers showed a greater change in an unfavorable direction on the Attitude Shift Index (subtracting January difference score from October difference score). Changers had higher Gross Change Weight scores, indicating greater exposure to unfavorable com- munications toward preferred store in October. Changers tended to leave the A & P more than other stores. Changers tended to be attracted to Kroger. 212 ll ma fl NH . a s a! HH 8 a s t! CH e II m e s tl m e s s. e :1 h « II o e e s e * rl m s a e a I! w s s e x II M L1 m 8 TI H e e s s a e s s e a 0 mm mm Hm ON ma ma 5H ma ma va ma NH Ha OH m m h o m g m N H mmqm m0 ZOHBéADmHmMQ MUmemmMHQ BZfiUHmHZOHm m0 NMdEZDm m.® mnmdfi 213 1. New residents v. old residents. As expected, new residents tended to be younger and have fewer children than older residents. 2. Life-cycle categories. No differences found other than on the old resident--new resident dimension. 3. Socio—economic categories. The respondents who had completed college tended to have lower predispositions ("A") toward their preferred store than those with some or no college education. The college graduates tended to prefer National and A & P. Higher status respondents also preferred National, but not A & P. The lower—status, lower-income, and less-educated respondents preferred Wrigley and Kroger. 4. Newspaper readers v. non-newspaper readers. It will be recalled that the sample included only 52 per cent who read the newspaper carrying the bulk of the advertising by super markets. This fact in itself was a biasing factor in that advertising pressure scores would be at a disadvantage when compared with face-to-face pressures to which nearly all the respondents were exposed. The persons with advertising pres- sure scores unfavorable to their October choice were predominant in the newspgper readership group. Newspaper readers also tended to rate advertising as more important source of infor— mation about super markets than ”self" or face-to—face. Among the media, the newspaper readers rated newgpapers and direct 214 mail as the more important sources. They also were the ones composing most of the group who shopped at more than one store "last week." 5. Opinion leaders v. non-opinion leaders. As noted, there were more opinion leaders among the changers. There was a significant difference in patronage patterns among the major super markets. The A & P and Wrigley attracted a larger number of opinion leaders than did Kroger and National on both waves. Curiously, while it was Kroger who gained the most customers during the interim, five of the seven opinion leaders (there were 36 opinion leaders in sample) who patronized Kroger in October went to A & P and Wrigley. Only two opinion leaders switched to Kroger. A & P, which lost fourteen cus- tomers, had a net gain of one in the opinion leader category. Relative to store attributes, opinion leaders were less in— terested in location convenience, more interested in cleanli- ness and variety of merchandise. There was no difference on "price." 6. Gregariousness. Only one difference was found when the gregarious group was cross tabulated against the non— gregarious group. Those who were gregarious had advertising pressure scores unfavorable to the super market preferred in October. Gregariousness was not associated with opinion leadership or with face—to—face influence, contrary to expectations. 215 7. Predisposition "A." The customers of the A & P had significantly higher predisposition scores than did the cus- tomers of the other major super markets. The customers of National had much lower predispositions. This is largely a result of the fact that 60 per cent of the A & P customers had patronized it prior to moving to Spartan Village, as com— pared with only 12 per cent of National customers. The respon- dents with weak predispositions toward their preferred store in October had significantly higher Gross Change Weight scores. As previously noted, a greater number of those with weak pre— dispositions were among those who changed their store preference during the study. 8. Predisposition "B." The respondents who had low attitude preference scores toward their preferred store in October tended to have face-to-face communications unfavorable to that store. They also had higher Gross Change Weights; that is, the overall communication pressure was unfavorable. As noted, these people also tended to change their store preferences. 9. Face-to-face Pressure. Respondents with face—to— face scores which were unfavorable or only slightly favorable toward their preferred store in October tended to be among the changers, to have weak attitudes toward that store, and to have unfavorable Gross Change Weights. 216 10. Advertisipg Pressure. Those with unfavorable adver- tising pressure scores toward their preferred store in October tended to be among the changers, to be newspaper readers, to be gregarious, and to have unfavorable Gross Change Weights. In addition, they felt that ”self" was much more important as a source of information concerning super markets than did the respondents with favorable advertising pressure scores toward their preferred store. The latter rated the media as a more important source of super market information. 11. January Attitude Preference. Two significant dif- ferences in addition to (0) above were: (1) the fact that A & P customers tended to have larger mean difference scores in an unfavorable direction than did the customers of the other three stores and (2) those with the lowest preference scores had larger unfavorable attitudinal shifts. 12. Attitude Shift. The respondents with the largest shifts in attitude in an unfavorable direction tended to be changers. Other than (11) above, no other differences were found. 13. Gross Change Weight. The five significant differences found have been discussed above. 217 CORRELATIONAL FINDINGS The several hypotheses concerning the effect of communications upon attitudes and behavior assumed there would be a positive correlation among these variables. Statistical Procedure. The weights assigned to each individual on each scale were grouped into frequency distributions for the panel and index numbers assigned to the categories, as shown in Table 6.12. The primary reason for indexing was to simplify coding for machine analysis. The ordinality of the measures is apparent. The simplifi- cation obtained by grouping the panel and the reduction of the computational procedure was felt to out-weigh the dis- advantage of some information loss. The indices computed were considered to have face validity; e.g., each plainly indicates that it was measuring the underlying variable. Both the coefficient of correlation £_and multiple correlation 3 were used to measure the relationship between the variables. The use of correlations is acknowledged as not being in accordance with strict statistical requirements in that equal intervals are assumed for the ordinal scales. The assumption that the variables were normally distributed seemed reasonable, however; therefore, the risk of getting answers 218 that may not quite correspond to reality was considered preferable to the loss of information entailed by restricting the analysis to statistics appropriate to ordinal scales. The calculations were made possible by the use of the integral computer at Michigan State University (MISTIC). Linear Correlations. The linear correlations com— puted for the communications influences and the dependent variable, Attitude Shift, are shown in Table 6.9. TABLE 6.9 LINEAR CORRELATIONS BETWEEN ATTITUDE SHIFT AND THE COMMUNICATIONS VARIABLES r Gross Change Weight .10 Face—to—face influence .19* Advertising influence .ll Predisposition "A" -.04 Predisposition "B" —.26* *Significant at the .05 level. Only face—to-face influence, among the three com— munication pressures, showed a significant positive correla— tion. Predisposition "B" (not included in computing the Gross Change Weight) showed a significantly negative corre- lation. The Face-to-face ;_mildly confirms the findings by Katz and Lazarsfeld (1957) which suggested that personal 219 influence is the dominant communications influence. Correlations between the Communications Variables. Correlations were also computed between the communications variables, as shown in Table 6.10. TABLE 6.10 LINEAR CORRELATIONS BETWEEN THE COMMUNICATIONS VARIABLES r GCW F-F ADV "A" "B" Gross Change Weight -— Face-to-face influence .56* -- Advertising influence .53* .10 -- Predisposition "A" .58* .08 .12 -- Predisposition "B" .24* .23* .10 .10 -- *Significant beyond the .05 level. The fact that the Gross Change Weight was computed by adding the scores received on Face-to-face, Advertising, and Predisposition "A” was expected to result in the strong correlations obtained between these scores and the GCW. The nearly equal ;:§_indicate similar contributions to the GCW, suggesting that advertising has a greater effect upon actual patronage than the "two—step flow" hypothesis implies. Face- to-face influence is significant when related to attitude formation, as seen by the correlation of .23 with Predisposition "B." The fact that the individual communications components show no significant correlations among themselves indicates that they operate independently of each other. 220 Correlational Analysis of Sub—groupingg, To further examine the data, correlations were computed for two important sub-groups: changers and non-changers, and the major store groups. For purposes of analysis, the store groupings as they were constituted in October were used. The "non—changers" are those who still preferred the same store when interviewed in January, while the ”changers" were the October customers who had changed their preference to another store. Tables 6.10 and 6.11 show only the correlations found to be significant beyond the .05 level. In Table 6.10, the same patterns of correlation exist for the store and change categories as for the panel as a whole. However, the signi- ficant ;_obtained between the Face—to—face and Attitude Shift indices (Table 6.8) can now be seen to be the contribution of A & P customers. It is not significant for either the changer or non-changer groups. The negative correlation found between Predisposition "A" and Attitude Shift may be attributed to the Kroger and National customer groups. These were also the stores who lost the fewest customers during the interim period. The implication is that attitudes ggp_be changed regardless of the intensity with which they are held, at least for super markets. It is also noteworthy that the impact of advertising on the changers is greater than either face—to-face or 221 mwol who mm. Gav. Qmol :m: om. mm. mm. me. he. on. :4: mg. ms. mg. mm. >U< me. so. mm. mm. mm. me. so. mum BOO 3mm mum 300 nmd mum 300 £m< mum BOO £m¢ mlm 300 Smd mum BUD awe mum BOO Smfi him 300 £m¢ mummcmnolcoz mumgcmno muwmQM£UIcoz mummcmnu mummcmnolcoz muwmcmnolcoz mpgggggo HMCOHumZ Homonx >0H®HH3 m c 4 mmMOZ¢EUIZOZ QZ< mmmwz¢m0 Mm QMNHMOOMBdU mmDOmU mmSOBmDU mmOBm meZBmm wZOHfidqmmmOO BZflUHmHZOHm NH.6 mum6< we. me. me. me. am. om. we. mommuouumomm pnmflmz mmcmsu mmouw smafim wpsuflppfi him 300 de mlm BOO £m< him 300 £m< him 300 Smfi him ZOO Smd him 300 flmfi Hmcoeumz ngoux >macfluz m w d mHmUcMQUICOz mumgcmzu mQSOHU muoum mmfluogmumo wgsmnu WMMOZ¢EUIZOZ QZ¢ mmmwz¢mv megfimm QZ¢ mmDOMO mmSOBmDU mMOBm mezBmm mZOHfiflflmmmou BZflUHmHZOHm HH.© mqmde 222 predigpggitions, and lowest of the three types for the non- changers. This suggests that advertising may be the "trigger" which will cause the change in behavior. When each store group is analyzed by Change cate- gories, as shown in Table 6.11, the non-changers within each group show consistent patterns. The changers, however, reveal marked differences. For the A & P and Kroger customers, the influence of predisposition has weakened. These were the two stores that had benefited the most from the tendency on the part of many respondents to remain loyal to stores patron- ized before moving to Spartan Village. Results for National customers show that the changers tended to leave National pri- marily because of the advertising influence of its competitors. Because of the small g for the store change groupings, extremely high correlations were necessary before significance could be established. Future researchers would be well-advised to use a large enough sample so as to obtain sufficient fre- quencies for such sub-categories. They appear to provide fruitful insights for the explanation of both attitude forma- tion and patronage behavior. Beta Coefficients - The relative importance ofthe individual independent variables in a multiple correlation in determining the dependent variable may be determined through the use of the beta coefficient. 223 The coefficients of the multiple correlation regres- sion indicate the increase in the dependent variable resulting from a unit increase in the indicated independent variable. The multiple correlations found for the Attitude Shift Index and the Gross Change Weight were .23 and .99 respectively. However, the independent variables, being expressed in dif— ferent units, made a direct comparison impossible. The coef- ficients of multiple regression may be made comparable by dividing each variable by its own standard deviation. The beta coefficients thus are comparable measures and indicate the increase in the dependent variable from an increase in one standard deviation in each independent variable. For the Gross Change Weight, the following were obtained: R = .9874 Betas: Predisposition "A" = .58 Face-to-face = .49 Advertising = .49 The results are essentially the same as those given by g, all variables are of about equal importance. Predispo— sition is the largest contributor of the three by a slight margin. For the Attitude Shift Index, the following was obtained: 224 R = .2281 Betas: Predisposition "A" = —.04 Face-to-face = .20 Advertising = .10 Again the results are similar to those given by ;, Face—to-face pressure is the only independent variable showing a significant relationship. In summary, the findings established that the shift in attitudes had little relationship with communications indices as constructed, with the exception of a mild correla— tion with personal influence. Many factors could account for the lack of relationship, e.g., the use of parametric statistics for evaluating the ordinal communications pressure indices. The attitude difference scores may have been a reason, although similar results were obtained when only the mean evaluative scores for ”own" store were compared on each wave, instead of taking the other super markets into consideration. The arbi— trary weighting scheme, both within the various indices and between the indices, could be improved. Yet, it seemed more realistic to look for other causes to account for the strange behavior of the attitude index. Actual Patronage Behavior. First, it is instructive to see how well the indices conmmred with actual patronage behavior. Inasmuch as the respondents who had changed pref- erence between waves could be accurately ascertained, it provided 225 a valid method for checking the predictive powers of the indices. Not only did the panel design provide observations relative to the change in verbal attitudes as determined on the semantic differential, but also the actual changes in patronage which took place over the three-month period. It is the latter behavior which constitutes the acid test as far as the advertiser is concerned, as well as being a de facto demonstration of attitude. Therefore, the changers and non— changers were tested against the null hypothesis that the sets of obtained frequencies in each index differ no more from a set of expected frequencies than would be expected by chance alone. If the communication indices could predict these respondents actually changing stores, it would constitute evidence of effect. Significant chi—squares were found for all indices except Predisposition "B." As hypothesized, the changers had weaker predispositions, unfavorable face—to—face pressure, and unfavorable advertising pressure toward the super market they had previously preferred. The cumulative communications index, Gross Change Weight, was found significant beyond the .001 level. The Attitude Shift Index also was found significant beyond the .001 level. In a sense, the fact that both indices were significant despite their failure to correlate with each other 226 is paradoxical. Statistically, it is possible to obtain significant chi—squares while showing little correlation linearly. The reason could also be due to the fact that two indices are pointing to two different kinds of changers, as will be seen in the deviate analysis which follows shortly. Table 6.13 summarizes the distribution of scores for the changers and non—changers on each of the change variables. Only for Predisposition "B" was there no dif- ference between the groups. Deviate Analysis - The failure of the Gross Change Weight to yield a significant correlation with the dependent variable, Attitude Shift, despite the fact that both were significantly correlated with actual changes in store patronage, suggested a difference in the respondents themselves. From a scatter diagram of the entire g_(X axis = Gross Change Weight, Y axis = Attitude Shift Index) the forty-seven respondents who deviated the most from the regression line were separated from the rest of the sample. (Deviates were those with unfavorable attitude shifts but favorable Gross Change Weight scores, or favorable attitude shifts and unfavorable Gross Change Weight scores.) The two groups were tested against the null hypothesis of no difference across all classifications and indices (see Table 6.8), using the chi—square statistic (significant beyond .05 level). TABLE 6.13 PERCENTAGE OF CHANGERS AND NON-CHANGERS FOR MAJOR STORES ACCORDING TO SCORES RECEIVED ON THE CHANGE INDICES 227 Index and Scores [High (1) to Changers Non-changers Total Low (7) toward October Choice] (n=35) (n=llO) (n=l45) Attitude Shift Index:* 1 2.8 3.6 3.4 2 5.7 9.1 8.3 3 5.7 21.8 17.9 4 28.6 39.1 36.5 5 14.3 18.2 17.2 6 28.6 7.3 12.4 7 14.3 .9 4.1 Gross Change Weight:* 1 0.0 14.6 11.2 2 8.6 17.9 15.4 3 8.6 26.9 22.3 4 25.7 26.0 25.9 5 28.6 13.0 16.8 6 28.6 1.9 8.4 Predisposition ”A":* l 5.7 13.6 11.7 2 11.4 17.3 16.6 3 0.0 2.7 2.1 4 14.3 16.4 16.6 5 28.6 36.4 34.5 6 25.7 9.1 13.1 7 14.3 2.7 5.5 Face-to-face Pressure:* 1 11.4 17.3 15.9 2 5.7 23.6 19.3 3 11.4 16.4 15.2 4 20.0 29.1 26.9 5 28.6 11.8 15.9 6 17.1 0.9 4.8 7 5.7 0.9 2.1 Advertising Pressure:* 1 0.0 6.4 4.8 2 5.7 15.5 13.1 3 14.3 19.1 17.9 4 28.6 32.7 31.7 5 11.4 12.7 12.4 6 22.9 12.7 15.2 7 17.1 0.9 4.7 228 TABLE 6.13. (Continued) Index and Scores [High (1) to Changers Non—changers Total Low (7) toward October Choice] (n=35) (n=110) (n=l45) Predisposition "B": l 5.7 0.9 2.1 2 0.0 8.2 6.2 3 14.3 21.8 20.0 4 22.9 29.1 27.6 5 22.9 16.4 17.9 6 22.9 17.3 18.6 7 11.4 6.4 7.6 *Chi-square between changers and non-changers significant at .05 level (six degrees of freedom for all indices except GCW, where df = 5). Significant differences were found as follows: Between major store customer groups - Kroger contained the fewest, A & P the greatest number of deviates, on both waves. Store attributes — the deviates were signifi~ cantly more price oriented. Stated importance of advertising, face—to- face, and "self" influence - the deviates rated face-to-face very low, advertising moderately low, and "self" very hign. Media importance — the deviates tended to rate newspapers very low in importance and the other media much higher than the non-deviates. Among changers: all deviates were satisfied with store they had been patronizing at time of change, compared to 60 cer cent of the non—deviates. No significant difference was found on any other cross tabulation. The pattern of the significances suggests the deviates tend to make decisions more on what Katz (1960, p. 177) 229 terms "utilitarian" attitudes: "Utilitarian needs are rein- forced by experience and not verbal appeals." The attitude shifts of the non-deviates show a high relationship with com- munication influences, suggesting different motivational bases. Their attitudes may be more the result of cognitive needs. It also suggests that the deviates are less persuasible. The deviate group itself divided into two nearly equal subgroups which were compared: HI—LO (highly favorable attitude shift toward preferred store in October but very unfavorable communication pressures, as shown by Gross Change Weight index) and LO—HI (unfavorable attitude shift but very favorable communication pressures toward preferred store in October). The g_for HI-LO group was 22; the p for the LO-HI group was 20. Significant differences were found as follows: Changers v. non-changers - all nine changers in the deviate group were in the HI-LO subgroup. Predisposition "B" — the HI-LO deviates had higher attitude preference scores toward chosen store in October. Media importance - the HI-LO group rated the broadcast media as the most important source of information about super markets among the mass media. The LO-HI group rated newspapers and direct mail as more important. (There was little opportunitytx>receive broadcast messages as the bulk of super market adver— tising appeared in the print media.) 230 Predisposition "A," face—to-face pressure, and advertising pressure - the HI-LO group had significantly lower scores on each of the communication indices which composed the Gross Change Weight, as well as a signifi— cantly higher score on the Attitude Shift index. It would appear that the HI—LO group simply are not print—media oriented relative to perceiving super market advertising. Inspection of the advertising pressure scores (see Table 6.4) revealed most of this category had neutral or slightly unfavorable scores, indicating little or no interest in the super market ads appearing in the newspapers. The fact that one-half of the HI—LO group was composed of changers suggests that these people shift preferences for utilitarian reasons. As one of these stated, "I still like the A & P, but go to Kroger because I'm saving stamps." It would seem that the deviates in the LO-HI category were still selectively perceiving advertising and face-to—face messages for their preferred store, for which strong predis— positions were previously held. The unfavorable shift in attitudes indicates dissatisfaction, but not enough to over— come habit. No differences were found on any other classification or index. In view of the slim evidence from this analysis, judgment as to causal factors should be withheld. The studies by Hovland, Janis, et a1. (1959), on personality and 231 persuasibility suggest that personality traits should be investigated. For example, persons who have low self-esteem or are of the female sex tend to be more persuasible. 232 CHAPTER VII DISCUSSION This study was designed to investigate the relation- ship of certain communication factors to super market choice in a natural situation. Instruments were developed to describe the characteristics, super market shopping and communication patterns, and attitudes of the representative panel of con- sumers; to measure changes in their attitudes and behavior concerning alternative super markets available to them; and to measure the communication influences to which the panel was exposed. Specific hypotheses were tested. Two sets of implications will be discussed. Theoret- ical implications will be examined in the first section. The findings relative to each of the hypotheses will form the basis of the discussion. In the second section, practical implica— tions will be discussed: (1) the findings relative to super market shopping and communication patterns; (2) the findings relative to image; and (3) the findings relative to the two- stage attitude change hypothesis. The chapter concludes with suggestion for future research in this area. THEORETICAL IMPLICATIONS The study of advertising effect on consumer behavior was seen to relate to the larger question of communication effect 233 on human behavior. The study of attitude formation and at- titude change was held to be basic to an understanding of the process whereby advertising campaigns achieve their ob- jective of influencing consumer behavior. Advertising ap- pearing in the mass media was viewed as one of three princi— pal message types. The other two, face-to-face and "self" messages, used interpersonal and intrapersonal channels. All three were seen to act in combination in achieving effect in the form of reinforcing or changing attitudes concerning the related message-object, or in changing consumer behavior to- ward the object. The messages, in turn, are mediated by the consumer's social environment and personality. The situation chosen, which met the conditions set forth as necessary for investigating the salient variables, was the everyday process in which the consumer must decide upon and patroniZe a super market at periodic intervals. The patronage act was viewed as part of a process in which the relevant attitudes were generated which determined the patronage act. The consumer's image of a super market was regarded as the evaluative, or attitudinal dimension of her meaning for the store. Her attitude toward the store was defined in terms of her score on the semantic differential. 234 It was suggested that the consumer's position on an attitude continuum concerning a super market would move in direction and intensity over time as a function of the combination of advertising, face-to-face, and predisposition- al messages related to the store. The technique for assessing the influence of the communication variables was developed especially for this study. Evaluation against the usual criteria for measuring instruments suggests that it has promise relative to its objectivity, comparability, and utility. It had face validity in that results obtained corresponded with results which obtained corresponded with results which common sense would suggest. Its principal limitations are apparent relative to reliability and sensitivity, particularly in the measure— ment of face-to-face communication. The instrument for measuring attitudes and assess— ing attitudinal changes was an adaptation of Osgood's se— mantic differential. Evidence was cited supporting the use of the instrument as one that met the criteria for measuring instruments, and for the use of the evaluative scales to provide an operational definition of attitude. Communication effects were viewed as changes in the meaning of the super market concepts along one dimension only: changes in the location of the concepts in the evaluative dimension provided by the semantic differential. 235 The specific hypotheses tested were derived from four general propositions. Each of the hypotheses will be discussed in turn: Hypothesis 1. A change in super market patronage is a function of the combined weights of communication unfavorable to a preferred super market, as measured by indices of "predisposition A," face-to-face, and advertis- ing messages perceived by the consumer. Hypothesis 2. The position of a consumer on an attitude continuum concerning super market X, as measured by the semantic differential, will move in direction and intensity as a function of the combination of messages and predispositions related to the super market, as in- dexed by the "gross change weight" scale. Hypothggis 3. A change in super market patronage is a function of the intensity of the preferential attitu- dinal shift in an unfavorable direction between a preferred super market and alternative super markets, as measured by the semantic differential. Hypothesis 4. Face-to—face messages will have a greater effect upon attitudinal shifts on a continuum con- cerning a preferred store than will advertising messages in the mass media. Hypothesis 5. The positions of a consumer on an attitude continuum concerning each of several alternative super markets and her "ideal" super market will reveal that the positions for the "ideal" and the super market preferred will be the most closely related. Hypothesis 6. The positions of a consumer on an attitude continuum concerning each of several alternative super markets, as measured by the semantic differential, will Show greater changes in direction and intensity fol- lowing a change in behavior, defined as a shift in patron- age, than will the positions of a consumer who has not shifted her patronage. Hypothesis 7. The proportion of residents residing on the same floor of an apartment dwelling patronizing the same store will tend to increase over time. 236 Hypothesis 1 was supported. The findings clearly showed the summation of the advertising, face—to-face, and predispositional message scores, as indicated by the Gross Change Weight Index, to be more unfavorable to the store previously preferred for those respondents who had changed super market preference. In addition, each of the separate message indices was shown to be more unfavorable for changers. The second hypothesis, stating that attitudes would shift as a function of the combination of messages and pre- disposition, was not supported. The failure of the Gross Change Weight Index to yield a significant correlation with attitudinal shifts, despite the fact both were significant- ly correlated with actual changes in store behavior suggested imperfections in methodology or the existence of influences operating independently of the communication variables. It was reasonable to assume that changes in patron- age constituted de facto demonstrations of attitude change. The lack of sensitivity of the indices and the analytical design may have been a factor in the low correlations ob- served between the attitudinal shifts and the communication indices. Yet, it seemed more realistic to look for other causes to account for the behavior of the attitude index. We had known from our study of shopping patterns that past experience was a powerful factor in the selection 237 of a super market. The predispositional index, however, showed no correlation with attitude shifts. Advertising was not related. Only face—to-face communication was re- vealed to have a barely significant relationship (mildly confirming findings from other studies supporting the notion that interpersonal message influence is more effec- tive than the mass media). Because each of these indices, singly and in combination, was strongly related to changes in super market patronage, a reappraisal of the data was made to see if any other factors could be isolated which might ex— plain why the indices did not correlate with attitudinal shifts. The ultimate practical test of the validity of an attitudinal measuring instrument lies in the usefulness of the measurements for the understanding and prediction of the consumer's behavior. If a consumer's attitudes toward several competitive stores are correctly assayed, it should be possible to make accurate predictions about how she will behave toward them. The findings on images generally sug- gested that we were at least partially successful in meeting this test. While the correlations found in the analysis of the indices used in testing the two—stage attitude change 238 hypothesis proved to be lower than desired, it might be hoted that this has been the general case in studies of attitude change (Katz, 1960, p.200). The overall organi- zation of attitudes and values in the personality is highly differentiated. In addition, the generalization of attitude change is limited by the lack of systematic forces in the situation necessary to implement the change. For example, the woman who prefers a more distant super market but does not have an automobile, so walks to a closer store, may still retain favorable attitudes toward the preferred store. The pattern of findings when deviates were compared with non-deviates pointed to the existence of a consumer type who is "more rational than others." The deviates were more price-oriented; they preferred the A & P (which had the low price image) over other super markets; they rated "self" as the most important source of information about super markets and face-to-face communication very low; they paid less attention to the print media, where the bulk of super market advertising appeared, than did the non-deviates. All deviates among the changers had expressed satisfaction with their previous choice at the time of change. One factor which the study did not investigate was change in knowledge resulting from exposure to messages in 239 the three major channels. Attitudes include both the affective dimension of liking or disliking, and also the cognitive, or belief, elements which describe the objects of the attitude. Specific attitudes are also organized into hierarchical structures, or value systems. Thus a generalized attitude toward a super market includes specific beliefs and feelings about the store's prices, quality of merchandise, etc. The semantic differential dealt primarily with the affective component; thus we were not making any allowance for changes in beliefs or the degree in differentiation between beliefs. The consumer is capable of acting in- telligently, and her decisions may or may not be related to predispositions and incoming communication. The as- sumption, for example, that exposure to an ad of a com— petitive store would be an unfavorable message did not allow for the fact that it might be perceived by the con- sumer so as to reinforce her beliefs about her own store. Findings by Festinger (1957, p. 176) indicate people may do so to avoid cognitive dissonance. Knowledge, as verified belief, will usually change when the object is perceived as changing; relevant attitudes will also change as a result. For the "rational” consumer, it could be surmised that each patronage experience required 240 verification and reevaluation of the super market in line with her beliefs. She either changed her attitudes toward the super market or reinforced them as a result. It was suggested that the deviates might tend to make their decisions on the basis of "utilitarian" attitudes, which is compatible with the concept of rationality. The non-deviates, on the other hand, showed a definite relation- ship between communication influences and attitudinal shifts. Thus their attitudes might have different motivational bases, particularly those serving their need for a cognitive structure. The nine changers among the deviates were all in the HI—Lo subcategory; that is, they had favorable shifts in attitudes toward the store they had previously preferred and an unfavorable communication change weight. This attach— ment to the former store suggested that the change in patron— age was temporary. The LO-HI deviates, on the other hand, had no chang- ers in their group. They were perhaps experiencing dissatis— faction with their preferred store for one reason or another, but still selectively exposing themselves to messages about that store for purposes of information or to avoid further dissonance. It is plausible to think of these people of changing patronage, at which time they will begin exposing themselves to communications about that store. Selective 241 exposure was clearly operant in the case of face-to-face messages, where four out of five respondents reported ex- posure to messages favorable to their preferred store. The influence of current experience is also seen in the breakdown by store customer subgroups across factors stated as important in super market selection. For the majority, it seemed that the choice was made on the basis of normative expectations. When the store was seen as not living up to these expectations, the respondent did not change her knowledge about the store, but rather the relative importance of her prior expectations so that they would coincide with what her experience verified. This was seen in the changes occurring in Table 4.14. Factors perceived as important in a store in October had shifted significantly for all super markets except the A & P. The deviates and non-deviates could also be thought of as non—persuasible and persuasible. There appears to be a general personality characteristic of persuasibility which has been found to be related to personality traits, such as low-esteem. Katz (1960, p. 204) suggests that such suscepti- bility to influence may help account for attitude change, but that it probably is not important. The finding that the changers were generally satis- fied with their previously preferred stores lends support 242 to Katona (1960, p. 121), who contends that highly moti— vated behaVior may arise without dissatisfaction, and that economic behavior does not support the deprivation theory of motivation. The third hypothesis, stating that a change in super market patronage is a function of the intensity of the preferential attitudinal shift in an unfavorable direction relative to the preferred store, was confirmed. The mean attitudinal shift of the changers was greater than for non-changers. Hypothesis Four was a corollary of the second hypothesis, and therefore subject to the implications dis- cussed. The finding supported the hypothesis. Face-to- face communications were revealed to be the only message component which was related to attitudinal shifts. The correlation was barely significant, but it does provide a mild confirmation of the studies supporting interpersonal influence as more effective in changing attitudes than the mass media. It is pointed out, however, that the analysis by customer groups showed the correlation to be a function of the A & P group. Among changers, only those leaving the A & P had changes in attitude which were related to face- to—face influence. Therefore, the finding that face-to- face communication is more influential than advertising as 243 it relates to store choice was considered by the author to need further investigation. Hypothesis Five and Six concerned the images people hold about an economic object. The findings supported both hypotheses: (l) a consumer's image of her preferred super market was closer to her image of an "ideal" super market than were her images of alternative stores; and (2) the consumer who did not shift her patronage during the study period tended to show a more stable relationship in the images she held of alternative stores than did the consumer who changed preference. The findings revealed the attitudinal patterns of the panel were quite stable over time although discernible trends could be noted. The most striking finding from the profile analysis was the fact that store customers had such a strong favorable attitude differential for their preferred stores on the clearly evaluative scales. Even in the case of the close/distant scale, the customers of the A & P "saw" the A & P as being as close as the shopping center stores, despite the fact the A & P was a mile further away. In other words, when choosing among relatively identical products, the consumer may make her choice as a result of her predispositions and exposure to face-to-face and advertising messages; then, if her experience with that 244 store is congruent to her expectations, she attaches the set of desirable attributes to her choice. When she makes a change to a competitive product, the entire cluster of attitudes will tend to shift so as to conform with the decision. The phenomenon suggests that, in the case of super markets, behavior may be easier to change than attitudes, thus lending support to the findings by Festinger (1957) and Kelman (1953). In other words, where the decision con- cerning an economic object is one that is sufficiently deliberative, it may be easier to take advantage of existing attitudes to produce a behavior which, in turn, will produce changes in attitudes. We saw in this study that changers invariably attached the entire cluster of favorable attitudes to their new choices. Yet the reasons for changing were not due to unfavorable experiences at the store formerly pre— ferred, but rather a ”pull" toward the new store. With Kroger, for example, whose advertisements had been judged to project friendliness and cleanliness (and the only ones other than the A & P's to appear in the campus newspaper) every one of the other major store groups saw Kroger as closest to their own store on these scales. In October, only the A & P store customers had done so. 245 It could be surmised that their advertising was an important factor in the Kroger success in attracting the greatest share of changers. The situation is not unlike what Mayer (1958) calls "added value." Given comparable products, if a benefit is promised concerning a product which is salient to the consumer's expectations, and the benefit is believed, the consumer will attach other desir- able features to the product as well, changing her attitude structure concerning the product. The last hypothesis, stating that floor groupings would tend to choose the same stores, was supported in the study. The findings relative to store influence agree with the considerable evidence supporting the existence of group influence, and suggest that neighborhood groupings may give rise to "group norms" which will affect the super market preferences of some members. PRACTICAL IMPLICATIONS Retail advertisers might be said to have three major "blind spots" overall: (1) concentration of short- term objectives at the expense of attaining long-range goals, (2) over-emphasis on price appeals, only one of many patronage factors, and (3) lack of adequate research (other than measuring advertising effectiveness in terms of immediate sales). 246 The principal task of this study was to provide a theoretical rationale for and a description of consumer attitudes and behavior in a careful and reliable manner. It is management's task to apply the results. Several of the findings appear to have value for merchandising and promotional application. The findings from the study did indicate advertis- ing to be a powerful determinant in consumer choice of super markets. The results indicated the value of estab- lishing long-range advertising objectives. It takes time to develop favorable attitudes among potential customers. Favorable attitudes of present customers should be rein— forced over time. Price was shown to be only one of several attribute variables influencing choice among equivalent alternatives (in this study the A & P should reinforce its low price image, the other stores might emphasize con- venience, friendliness, or cleanliness in their copy platforms). Shopping and Communication Patterns. While the . chapter on characteristics, communication and shopping paTl terns was primarily for the purpose of providing a des- cription of the ongoing process of super market patronage in a natural setting, the findings did yield insights which could be helpful in management decision-making. The method- ology also proved to have utility in developing data concerning consumer behavior. 247 The findings relative to the major store attributes which the panel stated were important in choosing a super market closely resembled those found in other studies. However, the reliance that management often places upon such generalized lists of store factors, usually rank- ordered in importance when reported by poll—type surveys, must be questioned. When the factors were analyzed by store customer groups, the pattern of responses was found to vary con- siderably over time. Each store group tended to develop a rank-ordering of specific factors which reflected different motivational bases. Changers proved to be generally satis- fied with previously preferred stores. Their shifts in behavior appeared to be due to other influences. The factors rated as important seemed to reflect social norms and values which the consumer selectively perceives in the store she prefers, especially when the alternative stores are ob- jectively similar. This is not to suggest that the con- sumer is not keenly interested in store attributes; she is cognizant of obvious differences, for example, between the major stores and minor stores in this study. Our analysis suggested that the majority of the panel tended to change the rank order of attribute impor- tance to correspond with the attributes they believed their 248 chosen store had or did not have relative to alternative stores. The finding thus can be restated as a hypothesis: super market attributes perceived by the consumer as im- portant result from patronage and are not a primary cause of store preference when alternative stores are objectively similar. In both the stated factors and on the polar scales, other store attributes were found to be more influential than price, at least when the price differential between stores was objectively similar. The finding that new residents in a community tend to investigate only two or three super markets within a short time after arriving before developing preferences which then tend to remain relatively stable suggests that the advertiser must seek ways to communicate quickly and specifically to them. For example, such media as direct mail, store signs, and the "Welcome Wagon,” might be used to supplement newspaper advertising so as to achieve more effective exposure for this important group. For the majority, super market shopping was a principal activity; one that was seldom done in conjunction with other shopping activities. This finding was supported by the fact the A & P was not located near any other stores, yet had little difficulty in attracting patronage. This 249 suggests that super markets locating in shopping centers may not benefit very much from the availability of other stores. As expected, consumers tended to selectively perceive communication favorable to their preferred store, as evidenced by the face-to-face and advertising pressure indices. Yet in the pattern findings, face—to-face con- versations were reported to be favorable in four out of five instances, while the readership scores indicated competitive advertising to have almost equal exposure. The readership of a given ad depended largely upon the quality of the copy and layout, not on the day it appeared nor the position within a paper. The use of "long copy" ads is questionable in view of the low "coop read” scores. While the ratio of reading "own" store ads seemed to be consistently in a ratio of 5:4:3 across the Starch- type readership scores, the scores for readership of ads of other stores varied widely. This again was apparently a function of the attention and interest factors of the ads. It also suggests that a store's customers read the ads for informational purposes, without need for appealing to factors they already ''know" the store has. Therefore, appeals should be selected to attract the interests of a competitor's customers. The advertising approach used 250 by Kroger proved to be effective, judging from the higher Kroger readership scores and patronage shifts to Kroger. It is the ability of the advertising to attract permanent new customers which will create the most profitable results in the long-run. The analysis of the communication patterns revealed the gross audience (those that were exposed to a channel in a given time period) was about equal for each of the major mass media channels. The message audience (those who remembered seeing or hearing an ad in that period) varied considerably between and within channels. A factor making channel comparisons difficult was the scant use of broadcast media by the super markets. Based on the scant data obtained concerning the effect of the two stores using direct mail and the one store using the broadcast media consistently, it would appear greater use should be made of these media, assuming similar costs. With only a small proportion of the panel "reading most" of the newspaper ads, other channels should be utilized to reach them. Image. The image findings were perhaps the most impressive insofar as implications for the marketer is concerned. The concept of image, as developed in the study, appears to have the potential of generating valuable data for guiding management in taking more effective remedial 251 and reinforcing action in their communication and merchan- dising mix. The graphic portrayal of image data proved to be a very feasible method for analysis as well as a simple, effective way for presenting complex qualitative data for interpretation by management. Attitudinal trends can be discerned from the profiles. In advertising, for example, the advertising by Kroger seemed to have been at least in part responsible for the shift on the pleasant and friendly scales. That verbal appeals should correspond to a consumer's experience was seen in the shift in the Wrigley profiles on the clean/dirty scales before and after remodeling the store interior and correcting the problem. The stability of attitudes suggested the need for long-term goals in advertising. Attitudinal trends were discernible which could be imputed to advertising in the case of the Kroger store (friendly and clean). That verbal appeals must correspond to a consumer's experience, however, could be seen in the shift in the Wrigley profiles on the clean/dirty scale, before and after the store corrected their problem. The attitudinal profiles have the ability to locate a specific problem—area, suggest what should be done, and report when problem has been remedied. 252 The semantic differential instrument, overall, was revealed to have considerable value for consumer research. It proved to have many advantages in obtaining attitudinal data, among them ease in application, reliability, sensi- tivity, flexibility, ability to delineate among the complex of factors studied, and face validity. The profiles were shown capable of locating specific problem-areas, suggest what needed to be done, and show when the problem had been met. Through the use of factor analytical or similar techniques, tailored scales and concepts could be developed to fit the advertiser, product and market. Extraction of clusters showing underlying images appears to have utility. Researchers in applied areas can catalogue customer and product types which should be valuable in explaining and predicting consumer behavior. The methodology described is relatively inexpensive and readily adaptable to national as well as local studies. Concepts used in the study were the store names. Components of stores could easily be added; e.g., manager, meats, produce, etc., to more effectively define consumer attitudinal positions. Two-Stage Attitude Change. The findings on the "two-stage attitude change" can be summed up in Himmelstrand's 253 (1960) comment: "anyone making a survey of the correlation of verbal attitude measurement with behavior will arrive at the disheartening conclusion that these correlations are in most cases considerably lower than we desire." A preconception of the study was that behavior would be a function of the intensity and direction of a consumer's attitudes. The findings did not fully support the hypothesis. Yet the exploration into reasons for the low correlation provided valuable insights. On the one hand, it appeared that those in the HI-LO (favorable shift in attitudes despite unfavorable communication) deviate category, some 15 per cent of the panel, were those tend- ing to "think for themselves" and whose rationality may have been a function of considered judgment as a result of experience with a store. For example, Martineau reports findings from one grocery study indicating only about ten percent of women are intensely price—minded. It could also have been a function of their "persuasioflity" dimension—- a personality trait which needs further definition. In addition, the phenomenon was observed repeatedly in the image and pattern analyses where attitudes seemed to follow behavior. Four factors thought to have effect on consumer patronage behavior were evaluated in the study: opinion 254 leadership, gregariousness, life—cycle position and socio- economic status. There was a slight indication that life- cycle and social status influenced choice of stores, but not on attitude change or patronage change.~ Opinion lead— ership was found to be important, but in an unexpected direction. Katz and Lazarsfeld (1955) saw the opinion leader as a person who influenced the choice of a product. The author found that, for super market selection, the opinion leader was the person most likely to change patronage, and also to change to stores different from those to which the non-opinion leaders tended to change. In other words, the importance of the opinion leader in influencing consumer choice would appear to need further validation. (Our method for detecting opinion leaders was patterned closely after that developed by Katz and Lazarsfeld.) We found no relationship between gregariousness and attitudes or behavior. Neither was this variable related to opinion leadership. The use of this trait as indexed in this study, seems to have little utility. SUGGESTIONS FOR FURTHER RESEARCH The findings of this study, in a sense, could be recast as hypotheses for further testing. The relation- ships between communication, attitudes, and consumer 255 behavior show considerable promise in explaining and pre- dicting consumer behavior. More information is needed in relating attitudes to behavior, and in defining factors influencing attitudinal change. The panel method provides an appropriate device for developing attitudinal indices over time which can be related to situational events. The semantic differential and factor analytical methods make the task of exploring attitudes feasible. Possible lines of future investigation in the area of communication and consumer behavior include the following: 1. The relationship between attitudinal shifts and behavior is perplexing. The relationship of other learning variables, particularly knowledge and skills, with attitudes should help explain the behavior of consumer over time. The testing of the notion that attitudes follow behavior should be fruitful. 2. The study of personality traits may be a source of explanation in accounting for behavior. The persuasible-—non-persuasible dimension appears to be a factor in determining advertising effect. Some consumers appear easier to influence than others regardless of the appeals used to change attitudes. 256 The semantic differential might be modified in order to obtain more sensitive measurements of competitive products. Including the potency and activity dimensions of meaning might be help- ful. Nouns, symbols, phrases could be evaluated for their ability to provide different shades of meaning than elicited by the bipolar adjec- tives used in this study. Such tailored scales could be useful in providing greater flexibility and suitability for the particular problems under investigation. The use of control concepts, such as the one used in this study, could be further explored. Major concepts could be broken into components for more refined analyses of at- titudes (Kroger manager, Kroger check-cashing, etc.). This would also help delineate the problem-areas better, thus aid management in taking more ap— propriate remedial action. The use of factor analytic methods for clustering variables and obtaining typologies of products, consumers, and messages would be helpful in marketing problem-solution. For example, our findings on the "shopping center" and "other store" clusters gave an indication of the direction of 257 patronage shifts. The "ideal" store cluster provided clues as to the relative importance of various store attributes. Mid-point judgments on the semantic differ- ential were perplexing. Did they mean the consumer had lack of meaning or was it the result of some other factor? For example, it appeared many respondents were reluctant to use the unfavorable positions on the clearly evaluative scales, but readily marked them for scales such as high class/low class. Inclusion of knowledge questions could clear up the relationship with concept familiarity. Expanding the scales by including the activity and potency dimensions of meaning would also be helpful. Relative to messages, the use of more objective techniques for obtaining exposure measurement (such as the Starch recognition method) need to be devised for interpersonal and broadcast media measurement. The indices used in the study were crude. In addition, the application of message measures should be done oftener, e.g., "last week," during the study period. 258 Content analysis of the media messages could profitably be used to see how they affect attitudes and behavior. The Kroger success in this study was felt to have been a func- tion of the structure and content of their ad— vertisements. Split-run techniques could have been used with the assistance of the advertisers to test effects relative to direct mail advertising. Investigation of reference group influence appears to be quite feasible in the natural situation as it relates to the buying act. The study design should include a provision for observing the behavior of all respondents in whatever groups are evaluated. For example, in the present study, had we extended our panel to include every person in the apartment build- ings by the simple expedient of asking them their preferred stores on each wave, much more information would have been available. At least two findings resulted from chance occurrences: the advertising by the A & P and Kroger appearing in the college newspaper, and the remodeling by Wrigley. By noting the positions of the consumer panel following the 259 first interviews and introducing purposeful changes in the situation through advertiser cooperation, several variables could be manipulated. Examples of these could be the use of different media, different content, remodeling of a store, special sales, "plant— ed" influentials, etc. 260 CHAPTER VIII SUMMARY AND CONCLUSIONS The purpose of this study was to investigate the effect of advertising on consumer attitudes and behavior toward an economic object. Advertising was seen as one of three major communication variables which function in combination with each other and whose effects are mediated by situational and personality factors concerning the message-object. The situation chosen was one where the variables operated in a natural setting and could be ob— served: the everyday process in which the consumer makes her decision to select, patronize, and change among the super markets available to her. Background. Empirical evidence accumulated to date relative to communication effect on attitudes and behavior point to the need for a better theoretical foundation to account for the anomalous and sometimes contradictory findings. From the existing literature, there is agreement that attitudes are predispositions to action and that a change in attitudes must be preceded by the reception and acceptance of a message suggesting a change. 261 Advertising, as a form of mass communication, is known to cause attitudinal change, usually in combination with interpersonal communication. How attitudes are learned is not clearly understood, nor is their relation to overt behavior. Theory. In the present study, the author envisioned that knowledge of the intensity and direction of the affec- tive component of a consumer's attitude toward an object such as a super market would enable him to predict the patronage act. The specific act of patronage was viewed as part of an evolutionary process in which relevant attitudes are generated which precede the act of buying. The two theoreti— cal positions underlying the study, images and two-stage attitude change, utilized attitude as the central concept for analyzing communication effects. Operationally, at— titudes were defined as the scores received on an adapta- tion of the Semantic Differential developed by Osgood and his associates, using only bipolar adjectival scales tested to elicit judgments corresponding to the evaluative dimen- sion of the meaning consumers had for the super market concepts associated with the scales. This dimension accounts for the largest share of the variance in the meaning people have for concepts. 262 The semantic differential thus provided data for describing store images as well as to test the hypotheses concerning the effect of the communication and other varia- bles salient to the study relative to changes in attitudes. Method. A representative sample of families was selected from the population of Spartan Village, a Michi- gan State University married housing development. The sample was interviewed twice, in October, 1960, and again in Jan- uary, 1961, thus becoming a panel. The data were Collected by means of a questionnaire administered by trained inter- viewers. Testing materials included the semantic differen- tial and newspapers carrying all the super market adVer- tising for measuring readership "last week." Results. Theresults were as hypothesized for six of the seven hypotheses: Hypothesis 1. Exposure to advertising and face- to—face messages, and predispositions were related to changes in store patronage. Hypothesis 3. Face-to-face messages tended to be more related to attitudinal shifts than were advertising messages. Hypothesis 4. The attitudinal profiles for the consumer's preferred super market and her 263 ”ideal" super market were more closely re- lated than those of alternative super markets. Hypothesis 5. The consumers changing store pre- ference showed greater attitudinal shifts for the several stores than did the non-changers. Hypothesis 6. Changes in super market patronage were related to the intensity of a consumer's attitudinal shift in an unfavorable direction. Hypothesis 7. The proportion of residents residing on the same floor of an apartment dwelling and patronizing the same store did increase over time. The one instance where the findings did not support the hypotheses was: Hypothesis 2. Attitudinal shifts were not related to the combination of messages and predisposi- tions concerning preferred super markets. The failure Of the attitudinal shifts to relate with the combined communication weights was explored and several implications drawn concerning the role of other factors. Conclusions. Theoretical and practical implications of the findings were discussed. Suggestions were made con- cerning further research. 264 The following conclusions are stated as courses of action that might be taken by researcher and practition- er, based on the author's interpretation of the findings of the study: A. Theoretical Findings 1. The findings revealed that changes in attitudes were related to changes in patronage behavior. Therefore, by study- ing attitudes directly as representing predispositions to action, it appears possible to obtain clues for the explana- tion and prediction of consumer behavior. The major communication variables studied-- advertising, face-to-face, and predisposi— tional or "self" messages--were found to be related, both singly and in combination, to patronage changes. The combined influence of the communication variables was not related to the attitudinal shifts. Face—to-face messages, while the only communication variable related (slight- ly—-mostly a function of the A & P customers) to attitudinal shifts, tended to have less 265 influence overall on patronage changes than did advertising. The evidence from the deviate analysis suggested that part of the reason for the failure of the attitudinal index to shift in the direction of the communication indices was due to personality factors. Some respondents tended to be more easily persuaded than others. We also found those who tended to be more price—oriented and tended to rate the importance of advertising and advice from friends very low and their self-experience high in evaluating stores. Further research on the general personality characteristic of per— suasibility would be desirable. The specificity and generality dimension of the belief component of attitude needs further definition. We saw that attitudes seemed to follow behavior in some cases, indicating the possibility of changing attitudes as suggested in Festinger's (1957) theory of cognitive dissonance. 266 Consumers hold significantly more favorable attitudes toward their preferred stores. The attitudinal profiles for the preferred stores closely resemble the profiles of the "ideal" store. For the aggregate, the attitudinal profiles remained quite stable over time, corroborating the finding that most consumers tended to select a unit of a chain store previously patronized when moving to a new community. The attitudinal profile changes did reflect the occurrence of events, how- ever, such as the remodeling of a store (Wrig- ley) and an advertising campaign (Kroger). There was little indication that demographic or socio-economic variables were related to super market patronage shifts. Contrary to the findings of Katz and Lazarsfeld (1955), there was little indication that the "opinion leader" influenced super market choice. The influence of the one type of reference group analyzed was positively related to patronage shifts. 267 B. Methodology 1. The design of this study provided a useful approach for the study of the processes of advertising and face-to-face effect and the direction of effect, in conjunction with consumer predispositions. By observing consumer behavior in a natural setting, the researcher can look for the specific act of patronage and assess the roles of the several influences which produced it. The concept of "image" and the methodology for defining it, as developed in this study appears to have generated valuable information for guiding management in taking more effective reinforcing and remedial action, such as a shift in ad- vertising strategy. The semantic dif- ferential revealed its ability to quantify the reactions of consumers to a super market. The attitudinal profiles indicated trends in the panel's reactions and relation- ships between situational events and consumer attitudes. The profiles also pointed out areas of strength and weakness of individual 268 stores. The graphic method seems an ap- propriate method for transmitting such complex information to management in sim- plified form. The typologies obtained through linkage analysis revealed its ability to cluster variables into types rapidly and objectively, whether for con— sumers, products, or items. Researchers in applied areas could eventually catalog customer and product types. Such a catalog would be a valuable aid in plan- ning strategy. C. Substantive Findings 1. The concentration of retailers upon ex- pediency and price appeals in their ad- vertising was shown to be questionable. Attitudes precipitating behavior changed very slowly over time. The findings con- cerning the influence of price, long considered by super market operators as the principal factor in obtaining increased sales, suggested that it was not an im- portant patronage determinant for stores other than the A & P. Thus it could be 269 concluded that the other stores had two principal alternatives available where effective action could be taken to in- crease sales: (1) the use of advertising and promotion designed to change or re— inforce consumer attitudes and beliefs concerning favored attributes other than price; and (2) by changing unfavorable attributes of the store or by innovation to take advantage of current consumer values. On the one hand, we saw that the A & P had by far the most favorable price image although its prices were judged by a panel of experts to be slightly higher than some competitors; Kroger's success in building "pleasant-friendly” attitudes may have contributed to its patronage increase. On the other hand, the reaction of the panel on the clean/dirty scale to an actual event, a change in the interior of the Wrigley store, demonstrated the effect of non-verbal communication on attitudes and beliefs relative to the pro- duct itself. 270 The notion of "image" promises to become increasingly significant for the advertiser as products and services become increasing- ly alike and objectively comparable with one another. Even in the case of a complex product such as a super market, the princi- pal stores in many communities can be con- sidered to be essentially equivalent choices. The store which is most successful in develop- ing a favorable image may obtain a competitive advantage by the careful nurturing of image components by changing related store attri- butes or by advertising and promotional cam- paigns designed to change the consumer's attitudes and beliefs toward the store, or by combining these approaches. The patterns of communication and shopping behavior revealed that such information can be of value in management decision—making concerning marketing strategy and tactics. For example, the generalized lists of favored store attributes obtained from poll-type surveys should probably be interpreted with caution. In the aggregate, the principal 271 factors stated as most important in select— ing a super market resembled those reported by other studies and in approximately the same rank-order: price, quality of merchandise, location, etc. While knowledge of the relative importance of these factors may have utility when applied to a product class, e.g., super markets, they do not discriminate between equivalent super markets. Favored attributes seemed to influence choice only when ob- vious differences existed between stores. When the alternative stores are objectively similar, the consumer selectively perceives favored attributes in the store she prefers. The findings were restated as an hypothesis: attribute rankings result from patronage and are not a primary cause for store selec- tion when alternative stores are objectively similar. Another finding of interest was that, for the majority, super market shop— ping was a principal activity; one that was seldom done in conjunction with other shopping activities. This finding was supported by the fact that the A & P was not located near any 272 other stores, yet had little difficulty in attracting patronage. This suggests that super markets locating in shopping centers may not benefit very much from the avail— ability of other stores. Advertising was positively related to changes in consumer behavior. The amount of readership of newspaper advertisements appeared to be related to the content and presentation of advertisements. The use of other media to supplement the newspaper was indicated. The finding that new resi- dents in a community tend to investigate only two or three super markets within a short time after arriving before develop- ing preferences which then tend to remain relatively stable suggests that the ad- vertiser must seek ways to communicate quickly and specifically to them. For example, such media as direct mail, store signs, and the "Welcome Wagon," might be used to supplement newspaper advertising so as to achieve more effective exposure for this important group. 273 REFERENCES Allport, G. W. The historical background of modern social psychology. In Lindzey, G. (ed.), Handbook of social psychology, Vol. I. Reading, Mass.: Addison—wesley, 1954, pp. 3—56. / Arons, L. Store image, frequency of shopping, and television. Submitted for publication to Journal of Retailing by Television Bureau of Advertising, Inc. New York, February, 1961. Applebaum, W. and Carson, D. Supermarkets face the future. In Seelye, A. (ed.), Marketing in transition. New York: Harper, 1958, pp. 13-30. Barnet, Edward. Showdown in the market place. Harvard Businegs Review, July-August, 1956, p. 85. Berlo, D. The proceps of communication. New York: Holt, Rinehart and Winston, Inc., 1960. Bogart, L. The age of television. New York: Ungar Pub— lishing Co., 1956, p. 215. Boulding, K. E. The image. Ann Arbor, Mich.: University of Michigan Press, 1956. Burgoyne Grocery and Drug Index, Inc., Continuing report, 1960. Cincinnati, Ohio: Burgoyne Grocery and Drug Index Inc., 1960. Cantril, Hadley. The invasion from mars. Princeton: Princeton UniverSity Press, 1940. Caples, John. Tested advertisipg methods. New York: Harper, 1932. Cartwright, D. Some principles of mass persuasion: Selected findings of research on the sale of United States war bonds. Human Relations, 1949, 2, 253-267. Charvat, Frank. Supermarketing. New York: MacMillan Company, 1961, Ch. 3. 274 Clark, Dan E., and Associates, Inc. Top ten brandg. A personal-interview survey conducted for the Youngstown Vindicator. Youngstown, Ohio, 1961. /’ Clark, L. H. (ed.). Conggmer behavior. New York: Harper, 1958. Cooper, E. and Dinerman, H. Analysis of the film, "Don't Be A Sucker:" a study in communication. Public Opinion Quarterly, 1951, 15, pp. 243-264. Crawford, J. W. Advertising: communications for management. Boston: Allyn and Bacon, 1960. Deutschmann, P. The image of paper X. East Lansing: Inland Daily Press Association, 1960. (Mimeographed.) Deutschmann, P. J., and Pinner, F. A. A field investigation of the two-stage flow of communication. Paper read before the Association for Education in Journalism, Pennsylvania State University, August 30, 1960. (Mimeographed.) Donnahoe, A. S. A new direction for media research. An address to a group of advertising executives at a luncheon meeting in Chicago, January 17, 1961. (Printed by Richmond Times—Dispatch and the Richmond Newg Leader, Virginia.) Duffy, Ben. Advertising marketg and media. New York: Prentice-Hall, Inc., 1951, Ch. 17. Ehrlich, D., Guttman, I., Schonbach, P., & Mills, J. Postedecision exposure to relevant information. Journal of Abnormal and Social Psychology, 1957, E3, pp. 98-102. Fearing, F. Toward a psychological theory of human communi- cation. Journal of Personalipy, 1953, 2;, pp. 71-78. Festinger, L. A theory of cognitive dissonance. Evanston, Ill.: Row, Peterson and Company, 1957. ' Fisk, G. Consumer information channels. Unpublished Ph.D. dissertation, University of Pennsylvania, 1956. Foote, Cone and Belding. The story of brand X. New York, 1958. (Mimeographed.) 275 Gardner, B. and Levy, S. The product and the brand. Harvard Business Review, Vol. XXXIII, No. 2 (March-Aprii 1955) pp. 33-390 Glock, C. Projective technigues in consumer motivation re- search. A report on a panel disucssion arranged for the national conference of the American Marketing Association. Ann Arbor: the Foundation for Research on Human Behavior, 1955. Green, B. F. Attitude measurement. In Lindzey, G. (ed.), Handbook of social psychology. Reading, Mass.: Addison- Wesley, 1954. Pp. 335—369. Hilgard, E. R. Theories of learning. New York: Appleton— Century-Crofts, 1956. Himmelstrand, Ulf. Verbal attitudes and behavior. Public ,Qpinion Qparterly, 1960 b, g5, pp. 224-250. Home Testing Institute. McCall's food and grocery products diary study. New York: McCall Corp. 1956. ,fHovland, C. I. Effects of the mass media of communication. In Lindzey, G. (ed.), Handbook of social pgychology, Vol. II. Reading, Mass: Addison—Wesley, 1954, pp. 1062-1103. Hovland, C. I., Janis, I. L., and Kelley, H. H. Communi— cation and persuasion. New Haven: Yale University Press, 1953. Hovland, C. I., and Janis, I. L. (eds.). Personalityiand persuasibility. New Haven: Yale University Press, 1959. a Hyman, Herbert and Sheatsley, P. B. Some reasons why in- formation campaigns fail. In Maccoby, Newcomb, Hartley (eds.), Readings in social psychology. New York: Henry Holt and Co., 1958, pp. 164-173. ~Hyman, H. H., and Sheatsley, P. B. Some reasons why in- formation campaigns fail. Public Opinion Qparterly, 1947, A, pp. 413-423. Katona, G. The powerful consumer. New York: McGraw Hill, 1960. 276 Katz, D. Preface. Public Opinion Quarterly, 1960 a, g3, pp. i-iii. “Katz, D. The functional approach to the study of attitudes. Public Qpinion Quarterly, 1960 b, 23, pp. 163—204. Katz, E., and Lazarsfeld, P. F. Personal influence. Glencoe, Ill.: Free Press, 1955. Katz, E. The two-step flow of communication. In Schramm, W. (ed.), Mass communications. Urbana, Ill.: University of Illinois Press, 1960, pp. 346-365. Kelman, H. Attitude change as a function of response . restriction. Human Relations, 1953, 6“ pp. 185-214. “/ Klapper, J. T. What we know about the effects of mass communication: the brink of hope. Public Opinion Qparterly, 1957-58. 21. pp. 453-474. Krech, D., and Crutchfield, R. S. Theory andyproblemg of §9cial psychology. New YOrk: McGraw-Hill, 1948. Lazarsfeld, P. F. Audience research. In Berelson, B., and Janowitz, M. (eds.), Reader in public opinion and communication. Glencoe, Ill.: Free Press, 1953, pp. 337-346. Lazarsfeld, P. P., Berelson, B., and Gaudet, H. The people's choice. New York: Columbia University Press, 1948. VLazarsfeld, P. and Kornhauser, A. Analysis of consumer actions. In Lazarsfeld, P. and Rosenberg, M., Egg language of social research. Glencoe, Ill.: The Free Press, 1955, pp. 392-403. :4/ ‘Lazarsfeld,,P. and Merton, R. Mass communication, popular taste and organized social action. The communication of ideas. .New York: The Institute for Religious and Social Studies, 1948. Lippmann, W. Public opinion. New York: Harcourt, Brace and Co., 1922. Lipset, S. M., Lazarsfeld, P. F., Barton, A. H., and Linz, J. The psychology of voting; an analysis of political behavior. In Lindzey, G. (eds.), Handbook of social pgychology, Vol. II. Reading, Mass.: Addison—Wesley? 1954, pp. 1124—1175. 277 ./’l s s s Lucas, D. B., and Britt, S. H. Advertis1ng psychology and research. New York: McGraw-Hill, 1950. MacLean, M. S., and westley, B. A communication model for communication research. Journalism Quarterly, 1957, 34, pp. 36-48. p/ Martineau, P. Motivation in advertising. New York: McGraw- Hill, 1957. Mayer, M. Madison avenue, U.S.A. New York: Harper, 1958. McNelly, J. Meaning intensity as related to the readership of foreign news. Unpublished doctoral dissertation. Michigan State University, 1961. McQuitty, L. L. Elementary linkage analysis for isolating orthogonal and oblique types and typal relevancies. Educational and Psyghological Meagurement, 1957, 11, pp. 207-229. Merton, R. K. Mass pegguasion. New York: Harper and Brothers, 1946. The Milwaukee Journal. Consumer analysisy 1960 edition. Milwaukee: The Journal Company, 1960. Mindak, W. A. Measuring advertising effectiveness from a communications point of View. Unpublished Ph. D. dissertation, University of Illinois, 1955. Mindak, W. Fitting the semantic differential to the marketing problem. Journal of Marketing, April, 1961, 22, pp. 28-33. Morgan! J. A review of recent research on consumer behavior. In Clark, L. H. (ed.), Consumer behavior. New York: Harper, 1958. PP. 93—219. Nafziger, R. O., and White, D. M. (eds.), Introduction to mass communications research. Baton Rouge: Louisiana State University Press, 1958. / , National Association of Food Chains. NAFC Management Clinic on Merchandising, Sales Promotion, and Advertising. washington: N. A. F. C. 1958. 278 Osgood, C. E. Cognitive dynamics in the conduct of human affairs. Public Opinion Quarterly, 1960, g3, pp. 341—365. Osgood, C. E. Report on development and application of the semantic differential. Institute of Communications Research and Department of Psychology, University of Illinois. (Dittoed.) Osgood, C. E., Suci, G. J., and Tannenbaum, P. H. The measurement of meaning. Urbana, Ill.: University of Illinois Press, 1957. Peak, H. Attitude and motivation. In Jones, M. (ed.), Nebraska symposium on motivation, 1955. Lincoln, Neb.: University of Nebraska Press, 1955, pp. 149-189. Politz, A. The rochester study. Philadelphia: Curtis Publishing Company, 1960. Remmers, H. H. Introduction to gpinion and attitude measurement. New York: Harper, 1954. Rogers, E. M., and Beal, G. M. Reference group influence in the adoption of agricultural technology. A study supported by the Foundation for Research on Human Behavior as a part of Iowa Agricultural and Home Economics Experi— ment Station Project No. 1236. Ames, Iowa: Iowa State_ College Department of Economics and Sociology, 1958. Rokeach, M. The open and closed mind: Investigations into the nature of belief systems and personalipy systems. New York: Basic Books, 1960. Rosenberg, M. J. A structural theory of attitude dynamics. Public Opinion Quarterly, 1960, ggj pp. 319—340. Schramm, W. How communication works. In Schramm, W. (ed.), The procggs and effects of mass communication. Urbana, Ill.: University of Illinois Press, 1954. Schramm, W. (ed.). Mass communications. Urbana, 111.: University of Illinois Press, 1960. Schramm, W. (ed.). Mass Communications. Urbana, Ill.: University of Illinois Press, 1949. 279 Schramm, W. (ed.). Theyprocess and effects of mass communications. Urbana, Ill.: University of Illinois Press, 1954. Schwerin, H. S. Competitive preference-—does it predict consumer action? An address to the American Association for Public Opinion Research, May 6, 1960. (Mimeographed.) Scott, W. D. The psychology of advertising. New York: Dodd, Mead and Company, 1921. Senders, V. L. Measurement and statistics. New York: Oxford University Press, 1958. Smith, M. Brewster. The personal setting of public opinions: a study of attitudes toward Russia. Public Opinion Quarterly, 1947, 11, pp. 507—523. Shannon, C., and weaver, W. The mathematical theory of communication. Urbana: University of Illinois Press, 1949, p. 5. The True Look of the Super Market Industry 1958. Super Market Merchandising, April, 1959, pp. 81-92. Star, S. A., and Hughes, H. M. Report on an educational campaign: the Cincinnati plan for the United Nations. American Journal of Sociology, 1950, 55, pp. 389-400. weiss, W., and Fine, B. J. Opinion change as a function of some intrapersonal attributes of the communicatees. Journal of Abnormal and Social Psychology, 1955, pp. 246~ 253. wells, W. D. Readiness to buy. Paper read at AAPOR-WAPOR Annual Conference, May 6, 1960. (Mimeographed.) Wiebe, G. D. Mass communications. In Hartley, E. L., and Hartley, R. E., (eds.), Fundamentals of social_psychology. New York: Alfred A. Knopf, 1955, pp. 159-195. \/ Wiebe, G. D. Merchandising commodities and citizenship on television. Public Opinion Quarterly, 1951—52, 15, pp. 679—691.‘ 280 Whittier, Charles. Creative advertising. New York: Holt, 1955, Ch. 18. Whyte, William H., Jr. The web of word-of-mouth. Fortune, November, 1954. Williams, Robin. American society. New York: Alfred Knopf, 1956, p. 404. APPENDICES 282 mmoprZQ OWZHMHN NVWVWZUHX my mu SHHon roareHoz om mcwmwzwwxmam mmmfl bmpmwsm. Swag. mu fig WHommH pnebom wHOm. : . zmdwesmw ©\Hm n H\s SH. Seogwmwp mfimdo . . c m . GSHnuk u§§¥ >Q§x< ' Nk 3»ng h... 33‘ 74,1 >3 ?. A . o 4!, .. U31; .2592; 2 4 I. a #0.. a APPENDIX C WAVE I CUESTIONNAIRE 28h IBM Code 1 2 3 4 5 6 7 (8-10 blank) SPARTAN VILLAGE SUPER MARKET STUDY Communications Research Center and Department of Advertising Michigan State University Residence of Apartment: .Interview Date: Oct. Interview with (person responsible for food store choice) SECTION I TO BE FILLED OUT BY INTERVIEWER (questions & ad readership) SECTION II TO BE FILLED OUT BY RESPONDENT (semantic differential) Note: 1. 2. Interviewer instructions are capitalized, follow closely. Complete all questions unless otherwise indicated. Where reSpondent cannot answer, put reason where it can be readily seen by coders. For example, in 2.3, if respondent talked to someone and can't remember name, put "can't remember." If answer is declined, put NR (no reply). Wherever possible, obtain information that will be helpful, e.g., can't remember name, probe to see if neighbor, friend who is not neighbor, etc. Use back of pages to complete answers where room in insufficient in Space provided. Be sure to identify with question number, place "see back” in question space. National supermarket - when "National” appears, this is the Frandor store. In a few cases, the Grand River store is patronized, be sure to identify by writing "GR" after store name (National QR). You can expect the unexpected: when in doubt, use your best judgment or call ED 2-1511, Ext. 2043, where Mr. Atkin will be available to provide assistance. All information is to be held in confidence. f Leave column blank 11 12 13 14 15 16 '17 18 19 20 21 22 23 24 25 26 27 1.1 1.2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 Page 1 (285) PART I What was the name of your favorite super market before moving to Spartan Village? (WRITE NAME & CITY) Why did you like it? (PROBE BRIEFLY - "MOTHER SHOPPED THERE,“) What super markets have you visited since moving to Spartan Village? (PLACE ”X“ IN BLANK ‘EFORE EACH STORE VISITED) A & P _u_National (F) Other Wrigley Prince Bros. (WRITE—IN) Kroger Schmidt's Please rank 1, 2, 3, the super markets in which you prefer to do your ShOpping in this area. (FUR FIRST CHOICE STORE, MARK "1" IN BLAhK IN FRONT OF 81053: SEfoflD CHOICE I‘2", THIRD CHOICE 3) ___A & P National (F) ___Other ___Wrigley _____Pr1'nce n ____'_Kroger Schmidt H Who or what led you to patronize state is your first choice? (INSERT IST CHOICE) (PROBE IN DEPTH) , which you Did you hear someone talk about the store before going there? Yes No (IF YES) Who? Name (MARK A, B, c, D AFTER "w110?") "A" - neighbor Name ”B" - friend, not a neighbor "C" - relative Name "D” - other (explain) Did you see any of their ads in a newspaper before going there? yes no Did you hear their radio commercials before going there? yes no Did you see their commercials on television? yes no Did you receive any mail advertising from them? yes no IBM Col. leave blank 28 29 30 31 32 , 33 34 35 2.8 2.9 2.10 3.1 4.1 5.1 5.2 Page 2 (286) Did you go out looking around for a store? l-yes 2-no (CIRCLE) If yes, what made you stop at that store? (PROBE) Summing up now, what was the most important of these in causing you to go to that store? (PROBE) ‘- -u p‘ -- ~...—.- a“ - A“. n..._v—I ——-” --—-~.'Q-~D-'- How often do you usually buy groceries? (CIQULE pkg) 1-Once a week Z-Twice a week 3-0nce every two weeks 4-Other (specifY) On what days do you usually do your food shoypiug? (CIRQCE ONE) l-Monday A-Thursday 2-Tuesday S-Friday 3-Wednesday 6-Saturday 7-Wednesday & Fri ay 8-Wednesday & Saturday 9-Other (specify) Where did you do your food sh0pping last week? (CIRCLE STORE OR WRITE IN IF UNLISTED. IF PERSON SHOPPED IN MORE THAN ONE STORE, CIRCLE EACH ONE AND PLACE APPROXIMATE AMOUNT OF MONEY THEY SPENT IN EACH TO SHOW RELATIVE IMPORTANCE). 12A & P S-Prince 3-Kroger Other 2—Wrigley 4-National (F) 6-Schmidt If shOpped in more than one store, Why? (PROBE) (e.g. like meats at second store, pick up milk and bread, because it is convenient, etc.) Page 3 (287) (SHOW PERSON LIST BELOW) THEN ASK.QUESTION "Which factor do you consider most important in choosing a super market? Next (2nd).........5th." PLACE CODE NUMBER IN RANK BLANK. Most important (lst.) ___ l-low price 2-location convenience 3-variety of merchandise 4-qua1ity of produce S-quality of meats Next important (2nd.) Next important (3rd.) 6-c1eanliness Next important (4th.) ____ 7-friendliness 8-parking facilities Next important (5th.) ____ 9-stamps (SHOW PERSON LIST BELOW. THEN ASK QUESTION.) "Where do you get most of your information concerning where to buy groceries? Please rank 1, 2, 3, the three .03" important sources.” (PLACE NUMBERS TO THE LEFT FE GREEK OF IMPORTANCE.) Most important (lst.) l-look for mysei’ 2-conversatiou w.,$ ~-:glbors Next important (2nd.) 3-friends, not neighcors 4-spouse or cihsf relatives Least important (3rd.) S-Lansing Sta:y_;purnal 6-Mailing pimcts 7-radio 8-te1evision 9-other What newspapers do you read regularly? (CIRCLE) l-State Journal ONLY 3-Both 2-State News ONLY 4-Neither IF THE PERSON DOES NOT READ THE STATE JOURNAL, OMIT QUESTIONS 8.2, 8.3, and 8.4; If read the State Journal, how much of the super market advertising do you read? (CIRCLE) l-never look at food ads 3-read about half of the ads 2-skim a few of the ads 4-read most of the ads Is there any particular day that you look for food ads? (CIRCLE) l-Wednesday 4-Mon., Wed., & Thur. 7-No particular day 2-Thursday S-Mon., & Thur. 8-Other 3-Wed. & Thur. 6-Every day c Il- . r. I. .0. .v. .3. r . v I. ‘1 . . .. l. . l .( '- c _ .l . .1. I‘ o .3. . v .: vi . D. u I ..l o I . . a .ll. . . . . _ -.~—o.—~. -c..- ‘1... 9.1 {2 9.3 Page 4 (288) Within any particular food ad, how much of it do you usually read? (CIRCLE) l-Scan 3-read about half 2-read less than half 4-read most of ad Which radio stations do you listen to regularly? (CIRCLE) l-WJIM.only 3-Prefer WJIM, but also listen to WILS 2-WILS only 4-Prefer WILS, but also listen to WJIM S-Listen to other stations mostly 6-Do not listen to radio OMIT QUESTIONS 9.2-9.5 IF PERSON IS A NON-LISTENER OF RADIO About how many hours do you listen to radio each day? (INTERVIEWER CIRCLE NUMBER OF HOURS BELOW) Hours before 6 P.M. (about) 0 1 2 3 or more Hours after 6 P.M. (about) 0 l 2 3 or more Do you remember hearing any super market commercials on your radio this past week? l-yes 2 -no (CIRCLE) (IF NO, OMIT QUESTIONS 9.4 AND 9.5) If yes, for what stores? 19A & P S-Prince Bros. Z-Wrigley 6-Schmidt's 3-Kroger 7-Other 4-Nationa1 8-Don't remember store 9. 4 What do you remember hearing? (e.g., GOLD BELL STAMP JINGLE 10.1 FOR WRIGLEY) A Which television stations do you watch regularly; WJIMkTV- Channel 6 or WILX & WMSB-Channel 10? Which station do you prefer? (CIRCLE) , l-WJIMrTV only 4-Prefer WJIMPTV but also Watch WILx-TV 2-WILX-TV only S-Prefer WILX-TV but also watch WJIMPTV 3-WMSB only 6-Other only 7-Don't have television (IF THE PERSON DOES NOT WATCH TELEVISION, OMIT QUESTIONS 10.2 - 10.6) . . .‘ . . . ' '. ' ‘ v.” . . . . ’ ‘ .r . , . . . 3. ‘ - . .L n' 3 -. .> ,. r . f .1.- .' . ‘ ~ I '. . , , _ _. . - g- . .L. . . ‘ . - . . a ‘ v...“ ‘ . .' ' .u. v - ’ - A an I ' ‘ I c . . . ~. : ' . . . I . - _ ,‘...r._ . k. . ._ . _. . _ -..r d ' . ‘ ‘ .‘V . .- ' .4 - ‘ L- I 3. - - . .. . - > . . n . I « ‘ I a ' ' r . . . .'. . - - a g :' . ._.l .> , 7. 4' ' ‘ ' - . . . . as . . . I , .- . ., _ . . . . . . . ‘x _/ 1‘ \ J I t ' \ A. . ' .1 ~ ' o t . . , . . . ' . . ~ ‘ ‘ I l - . . - o , . . ..4~. . . . 1.. . f ‘_. . J. ‘ . _ .' ‘ I I . . . . . 3 . _ I . . . . n I ' -- a . . - .- - I . .1. ‘ . .- ,.. ‘v...~.-....-... .-.......... . . . ' . v ._ D " ‘ 4 I‘ ,b' A ‘ ~ . . -- J. I . .-‘ ., A A , I ' . ’. I '7 "I a I - I ' , .. . . 1 - . . A . - 1. ., ~ . .’ A A . v . -, ‘ - g o - ( . - ~ - ‘ O ' ' .. -. . ., , . . -' ' v 1 M f , ' ~.u....--—. -—_.—._—\-..~.... ... .-.- ..._ .. .._.-,. .. .. - .. . ..... . t _. . .- ,. . . . . . - , . ,_.,_ ..... _.,.. .. .... -_..........- ,, .. .. .... . . — . Hr... . . .. .. §»-£ .. . . . . ._ . , , , H.,. __ _.7 . . _,,__ «as».-. r.-- ....—.—.-- ...,‘ .. .. .. -. «c— 4. s4 . . . ._. . .... H . . . . . . ._ -, _ ,, .,_ ._ _, _ -- , G o ' . n v' I. ' ' ‘ I ' - . . - . 1 - .. . ‘, 4: . . ‘ . . . . . - ' Q -—.-~-- . ‘ - . . ' . 3 . - . ~ \ . 1 . ,.. i~ . g . . ‘ _ ‘ . l - .. . , . . l . .- . . . . - . .. \. . .. ~I . ' . , _ , ~. ' . . .. . 1. . . - ; . 1 . . . ' . 4‘ . . I . ' ’ . ~ I . ‘ .o:- ' . - ‘ A . cc . . . v .- . , t g. .‘ . - ~ . . . . _, n D 2 ‘ -~.. - .'_ , .. ~.~_- . .. -V- . - .-~-....... - ... . . . .-. . - . ' ‘ u , I _, . .. .-. - . '0‘ \ ‘ t ,. .' "‘ . ‘ ‘ ‘ ~ \ , . - - . l . . _ ' . , - , . ‘ a ‘I ' . ' . . . . : IBM COL. 54 55 56 57 $8 59 60 10.2 10.3 10.4 11.1 11.2 (289) Page 5 How many hours do you watch television each day? (INTERVIEWER: CIRCLE NUMBER OF HOURS BELOW) Hours before 6 P.M. (about) 0 l 2 3 or more hours Hours after 6 P.M. (about) 0 1 2 3 or more hours Do you remember seeing any super market commercials last week? yes no (IF NO, OMIT QUESTIONS 10.4 AND 10.5) If yes, for what stores? (CIRCLE) l-A & P S-Prince Bros.. 2-Wrig1ey 6~Schmidt's 3-Kroger 7-Other 4-Nationa1 8—Don't remember store(s) Do you remember what you saw? Please describe. Do you ever talk to anyone about where to buy food? (CIRCLE) l-never 2~sometimes 3-often (IF TALKS ABOUT FOOD STORES) Would you mind describing the persons with whom you have discussed food stores? (INTERVIEWER: GET NAMES, IF POSSIBLE, BUT BE SURE TO FIND OUT.IF.THEY ARE NEIGHBORS,“WOR- ASSOCIATES, RELATIVES, MEMBERS OF CHURCH GROUP,. OR- CLUB ,- ETC. IF NEIGHBOR, FIND our WHERE THE PERSONS) LIVE, E.G. , SAME FLOOR, NEXT APARTMENT BLOCK, ETC. PROBE TO SEE WHICH ONE GAVE ADVICE ON WHERE TO SHOP.) Talked to Talked to Talked to ~~~. mu och". _ - .. a.--.--- _. 0.... I'. ~.- - .4. -—-.-.~u.. ‘.., . o‘nnu—o.--.'~ -....-... _. .. 26;" ' I ’ . '. ‘l - : I: ‘ .l I , ‘ I,‘ - ~ .1 ' .1 "': I .'| ‘ 'l. :-\.. I' " “ v'" *1 . '.. I)” 0 113“ my . , .. ' . I . O " ‘ . ' ' n . . . . .. . . ..- .., . um . u- “M..." . an... . n- . ~-.- ., .A, .. ~ - .. . . ., .o o .g- .. .0... . - s. .h. .... . . u ‘. ..-... . . .... .« — .,,.n on....-. . ,1... .-o 0.. . .. Io ....... ...-..‘....A~ *Cv. . ...,.-.~. . - ‘ J u . -. . . ...-.. ' . . L . .t . . .‘ . . , I . .. , .7 , . . _ ' ' t _ . » . ,. . . ‘ f ‘ . 4 I .- a . Y I . I . n ' ‘ .. , . u . . \ " ' . . _ ‘ . . ~| -,_ Jr- . ' \ .l ' ' \ ‘ ‘ . I In ‘ ' » ‘ ,- . ‘ .. . . ..... _-.. - . ..,.,-_ . u l 4f . .e . - . . . . . I .. . . . ~ . — ., . . . . . . . ‘ » g~ '_ 't ,1 ‘.. u ' .- D- ,.. .. . ...... . , .... . . _ -—.. . . . .. ... .. . .... .... . . . .. . _ . I . . . . . ’l . . . . . , . I . I, r ‘, .. . l .- . ~ . ' A. A l . ' '- . . ,. .( '. 1... . .. . . A . . . . - \. . . _ r . .‘ _ - ‘ -,‘. | 1. . ‘ ~ I . . . , . . ' . P ' _ \ .,. L l . . . . .. -.. I .. . . . I . ‘ ’ - .y .-. l 1 ‘ '_ \ o_ I . , ‘o "l . a - P ._ . "I ‘ I . . . ' I u I . . ; ‘ . . . ‘. . 4 ’ . , ~ ~ ‘ ‘ —- y . -' c . v . - ' . A l. ‘.\.’ I, ' ‘..I. e. - A s I I ‘ I" .I . - : - . - .1 . . . u . . -v-..- -. - _ . - .. .. . ... , . ,. . . . .. . n. . . . .. .... . . -4.-. . .. .~ - I, .-.~...~ . . u... -. --p.1.. . . . . . . -.. L , . . . .‘ _ ,_... . ... . a." .. . . r n: . .. - 1.. . . .. m— r - .. .- I . ..L .. . . I P _‘ .; .. ‘ Cl .. u- — I. . a -: ~. —. .. l“ (: w“. -sn IBM COL. 6L___ 62__g 61__g 64__g 6S__g 66-791- - 80__l CARD 2 1* 2— 3- 4* 5_ 6— 7— 8-10 BLANK 11___ 1A__ 13___ 14___ 11___ 1q___ 12___ 1&___ 13___ 20 11.4 11.5 12.1 12.2 12.3 13.1 13.2 13.3 13.4 Page 6 (290) Do you remember talking about a food store with anyone last week? What store(s)? (PLACE AN "X" IN THE SPACE TO THE LEFT OF EACH STORE SHE TALKED ABOUT). A&P __“Rroger Other (specify) Wrigley National What was said? (PROBE) —.'I'-'I".‘Q.v‘ — —— «ow—r1. _—-—_ .fi, m.-“..-..--.._._._-‘-.- ‘— -- —.-—0 ' an“ .—m - Do you receive any grvccry advertising in the n.31; no you read it? (CIRCL3} l-do not receive any 3-yes, read it 3u“£ itan 2-yes, read it usually Q-ycs, never flcgg : Did you receive any last week? Frow Wan? 'u..u;*§ ”A" IN THE SPACE TO TEE LIFT OF EACH STORE Ik“fi.g“lfid SHE RECEIVED MAIL AD‘JELJIL‘IH‘J) . A & P Kruger Other Wrigley National .. m..- 3 _. 9 What did it say? (PROBE) Please answer the following data about yourself. Age. 1-under 24, 2-over 24 (CIRCLE) Number of children. l-None, 2-one, 3-two or more (CIRCLE) Last grade in school completed. (CIRCLE) 1- No college 2- some college 3- college graduate Do you hold a job or go to school? (CIRCLE) l-housewife 3-work only 2-student only 4-work and go to school IBM COL. 21 22 23 24 25 26 27 28 29 30 31 13.5 13.6 13.7 13.8 14-1 14.2 14.3 14.4 15.1 15.2 15.3 Page 7 (291) What is your father's occupation? (WRITE IN) Where are you from? (CIRCLE) l-Lansing area 2-Michigan, not Lansing 3-Out-of-state How long have you lived in Spartan Village? months How much do you spend on groceries each week approximately? (CIRCLE) l-Under $10, 2-$10-$15, 3-$l6-$20, 4-$21-$25, 5-$26 more Do you do your grocery shopping at the same time yru do another activity? (CIRCLE) l-rarely, Z-sometimes, 3-often, é-usn”lij What other activities? Describe (PROBE) m- . ‘I—l - J'I’ -w_ .1! 151‘. - - «qr—z...‘ _-.-\-- Do you usually go grocery shopping with any.we else? If yes, with whom? (CIRCLE) l-No 2-Yes, with someone in my family 3-Yes, with another person (outside my family) Who decides the store? Have you recently been asked for your advice on: a community problem? l-yes Z-no (CIRCLE) a personal problem? l-yes 2-no (CIRCLE) a grocery problem? l-yes 2-no (CIRCLE) To whom have you given your advice? (name and address) Compared with other persons belonging to your circle of friends, are you more or less likely than they to be asked for your advice on: a community problem? l-less 2-more (CIRCLE) a personal problem? l-less Z-more (CIRCLE) a grocery problem? l-less Z-more (CIRCLE) IBM CARD 2 COLUMN 35 36 37 38 39 4O 41 42 43 I I 44 45 I I 15.4 16.1 17.1 17.2 18.1 Tues. Wed. Thur. Page 8 (292) Do you like to meet people, go to social gatherings? (CIRCLE) 1--do 2--do not Do you know anyone whose judgment you would trust when it comes to getting advice on super markets? (DESCRIBE RELATIONSHIP - FRIEND IN CHURCH, NEIGHBOR, MOTHER, ETC. - AND PROBE BRIEFLY "WHY", AND IF THEY LIVE NEARBY) How many peOple are there with whom you are friendly and talk to fairly often? Retghbors: about (number) Other friends: How many clubs and organizations do you belong to? (NUMBER) Groups in Spartan Village Other Campus groups Off-campus groups (READERSHIP OF SUPER MARKET NEWSPAPER ADS - READERS OF STATE JOURNAL ONLY) I have here c0pies of last week's State Journal in which there were super market advertisements. Would you please look at each issue with me and tell me which ads you saw? (LEAF THRU EACH EDITION, STARTING WITH MONDAY. CHECK ONLY THE ADS LISTED BELOW. INSTRUCTION SHEET DESCRIBES HOW.) Ads Noted Seen-Associated Read Most Coop Shoptite (Wed. only) A & P Shoprite National A & P Kroger Wrigley Quality Bob's Mkt. Beeman's ShOprite Universal Junedale Shaheen's Schmidt's (293) Code PART II The last part of this questionnaire is designed to measure the meaning of different words by the use of the "rating scale" shown below. Study the example to see how the rating is done. Advertising healthy : : : : X : : : : sick By placing an "x" in one of the spaces you can show the direction and the intensity of your association with the word "advertising." For example, suppose you feel that "advertising" is neither "healthy" nor "sick". You would indicate your feeling by placing the "x" in the middle Space as shown above. However, should you feel that "advertising" is very "healthy" you would mark the scale like this: Advertising healthy : X : : : : : : : sick If your feeling is that "advertising" is guite "sick" you would mark the scale as follows: Advertising healthy : : : : : :X: : sick However, if you feel that "advertising" is. only slightly one way then you should check like this: Advertising healthy : : :X: : :___: sick Now, would you mind checking your rating for "My Ideal Supermarket." MY IDEAL STORE good :___:__:___:__:___:__:__: bad fair :___:_:___:___:___:__:___: unfair pleasant :___:__:___:_:___:___:__: unpleasant clean :___:__:__:__:___:__:__: dirty friendly :___:__:__:___:__:_:____: unfriendly close :__:____:___:___:___:___:____: distant expensive :_:_:___:___:__:_:__: inexpensive progressive :___:__:___:___:___:___:___: conservative high class : : : : : : : : low class ---NOW - TURN TO NEXT PAGE AND DO THE SAME FOR THE SUPER MARKETS LISTED. ---WORK AS QUICKLY AS YOU CAN. MOST PERSONS FINISH THIS ENTIRE PAGE IN LESS THAN FIVE MINUTES. ' ‘ "'PLEASE CHECK EVERY ITEM. (IF YOU ARE UNFAMILIAR WITH THE STORE, CHECK CENTER SPACE). .- .n a V r ..— KROGER good A & P fair SCHMIDT'S pleasant NATIONAL clean WRIGLEY friendly PRINCE BROS close A & P expensive SCHMIDT'S progressive NATIONAL high class WRIGLEY good PRINCE BROS fair KROGER pleasant SCHMIDT'S clean NATIONAL friendly WRIGLEY close PRINCE BROS expensive KROGER progressive A & P high class NATIONAL good WRIGLEY fair PRINCE BROS pleasant KROGER clean A & P friendly SCHMIDT'S close WRIGLEY expensive PRINCE BROS progressive KROGER high class A & P good SCHMIDT'S fair NATIONAL pleasant PRINCE BROS clean KROGER friendly A & P close SCHMIDT'S expensive NATIONAL progressive WRIGLEY high class KROGER fair A & P pleasant SCHMIDT'S friendly NATIONAL close WRIGLEY progressive PRINCE BROS high class A & P clean SCHMIDT'S good NATIONAL expensive WRIGLEY pleasant PRINCE BROS 'good KROGER close SCHMIDT'S high class NATIONAL fair WRIGLEY clean PRINCE BROS friendly KROGER expensive A E P progressive oo ’0' o. co O. O. O. O. O. O. O. O. O. 0. bad unfair unpleasant dirty unfriendly distant inexpensive conservative low class bad unfair unpleasant dirty unfriendly distant inexpensive conservative low class bad unfair unpleasant dirty unfriendly distant inexpensive conservative low class bad unfair unpleasant dirty unfriendly distant inexpensive conservative low class unfair unpleasant unfriendly distant conservative low class dirty bad - inexpensive unpleasant bad distant low class unfair dirty unfriendly inexpensive conservative (291.) .,-.- o . en-.. 77-...” . , an o , ’ .' .‘ ‘ .,‘. u u . - c ‘ .» . ~ i , v ., n I , . . ' I a -.-i . ".'~' . . c. a n c u ,’ -"v. ' I . , , .4 . . . I A > - . . A . , . APPENDIX C . WAVE II CUESTIONNAIRE a295 IBM.Code SRARTAN VILLAGE SUPER MARKET STUDY Communications Research Center and Department of.Advertising Michigan State university Residence of Apartment: 1 Interview Date: Interview with (person responsible for food store choice;— v ‘— SECTION I TO BE FILLED OUT BY INTERVIEWER (questions & ad readership) SECTION II 10 BE FILLED our srnssrounsm (semantic differential) Note: 1. 2. Interviewer instructions are capitalized, follow closely. Complete all questions unless otherwise indicated. Where respondent cannot answer, put reason where it can be readily seen by coders. For example, in 2.3, if respondent talked to someone and can‘t remember name, put "can't remember." If answer is declined, put NR (no reply). Wherever possible, obtain.information that will be helpful, e.g., can't remember name, probe to see if neighbor, friend who is not neighbor, etc. Use back of pages to complete answers where room.in insufficient in space provided. Be sure to identify with question number, place "see back" in question space. . ,National supermarket v when "National" appears, this is the Frandor store. In a few cases, the Grand River store is patronized, be sure to identify by writing "GR" after store name (National 9;). You can expect the unexpected: when in doubt, use your best judgment or call ED 2-1511, Ext. 2043, where Mr. Atkin will be available to provide assistance. All information is to be held in confidence. .o I x u I. COI- . A I . 0. e u . v u s . o u . . . I; u a e .1 r . ' . u ‘u . up f a a \i .2 a . \ .. (Leave blank) 12 13 i4____ 15 I. 17 is 19__ 20 21 22 23 24 25 26 1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.4 Page 2 (296) PART I What super markets have you visited since the beginning of the fall term? (PLACE "X" IN BLANK FOR EACH STORE VISITED) A & P Kroger Prince Bros.-Sh0prite Wrigley's National Schmidt's Other (WRITE IN) Which of these super markets do you prefer as your first choice? Second choice? Third choice? First choice Second choice Third Choice Where did you do your shopping last week? (IF PERSON SHOPPED MORE THAN ONE STORE, FIND OUT WHICH WAS FIRST, SECOND, ETC., RELATIVE TO AMOUNT OF MONEY SPENT AND WRITE l, 2, ETC. IN_) A & P Kroger Prince Brothers Wrigley's National Schmidt's Other Which factor do you consider the most important in selecting a super market? Which do you think is the next most important? (TO 5TH RANK--SHOW LIST BELOW TO RESPONDENT) Most important First ___. l-low price Next most important Second ___. 2-location convenience ‘0 0 Third ____ 3-variety of merchandise -- « Fourth ___. 4-quality of produce " .. Fifth ___, S-quality of meats 6-cleanliness 7-friendliness 8-parking facilities 9-stamps Have you recently been asked for your advice concerning super markets? Yes No Compared with other persons belonging to your circle of friends, are you more or less likely than they to be asked for your advice on a grocery problem? More Les 3 How many people are there with whomeou are quite friendly and talk to fairly often? (ESTIMATE, NOT EXACT) Neighbors Other friends How many clubs, organizations, formal groups do you actively participate in? About 27 28 29 3O 31 32__ 33 35 36 3.1 3.2 3.3 4.1 4.2 Page 3 (297) Listed below are three major sources of information from which you can learn about super markets. Which of the sources do you feel are the most helpful to you in making a decision as to which super market to patronize? A - Media (newspaper, radio, tv, mail) Most important B - Talking to other people C - Looking around for myself Second most " Third most " Among the various media, which have you found to be the most helpful in deciding where to buy your groceries? A - Newspapers First B - Radio Second C - Television Third D - Mail advertising Among the various people you talk to, which are the ones you talk with most about super markets? A - spouse or relatives First B - friends, not neighbors Second C - neighbors, same floor Third D - neighbors, other E - store employees, teachers, etc. Do you remember talking to anyone about (SPECIFY EACH STORE, ONE AT A TIME) last week? Please try and recall if what was said was favorable, unfavorable, or "mixed." (CIRCLE) A & P? ‘___‘Yes ____No (IF YES) Favorable Unfavorable Mixed Wrigley? __ Yes __ No Favorable Unfavorable Mixed Kroger? ___ Yes ___.No Favorable Unfavorable ‘Mixed National?___ Yes ___ No Favorable Unfavorable ‘Mixed Now, thinking over the past three months, can you recall talking to anyone about (SPECIFY EACH STORE)? Would you say the dis- cussion about (EACH STORE) was, in general, favorable, unfavorable, or mixed? A & P? ___, Yes ___ No (IF YES) Favorable Unfavorable Mixed Wrigley? Yes No Favorable Unfavorable Mixed Kroger? Yes No Favorable Unfavorable 'Mixed National? ___, Yes No Favorable Unfavorable ‘Mixed . a . . . .. 0 Rio. 0 F. . . . o . . u . . a . n. O n a a. . I . u .. ~ 0 . . n a v2 .. u a m a. . . .. r. ... . . .. \I . . I Y. n . ~ - . n. . o .. . .n . u a . . . . . II n . D . l n I . n n I V a I .. u . r u . . s . n J o I .1 w .. A I .K In .- \» a o I l as I .l u l l n . a. . . .. a .. . a I s u . I a l v . . .. . . n . . .O. . I u r n V O . . . . 5| .0 1 l.. ,5 . . '4 .o c . . a v . II . . . l . . I. . . ,v .. Ia . o. - u. .. l. . o . o . I. . . . . a . a 1 ~ . . . u.. I... I .lqi . I 1 vi 0.. r. vi~ . l ‘ u . .1 . f. I. II I. . .o. A c o .l u I. .. . Ila I C o I . ..a .J .1 a A e . c .u l ,I .- . .. - Ia|a I n . RI: A o u .I I. .o.. uh- . . . S. c.- . . n . Q . . . . p - . .. . . s . a I u I . .~o . . I O n .1. L I: . . q . .u. . a . I . ... Kl. ( o... I :I I . .. . .. .o o . . . . . 4 l . . a ., .. . .. . . . I. o . ... I nu .a an \ . ‘0- .2. .u. . I. . . . . 5.1 37__ 38_ 39__ 40_ 41__ 5 .2 What newspapers State Journal do you read usually? State News (IF READS EITHER OR BOTH NEWSPAPERS) Page 4 (298) Neither Do you remember seeing any newspaper advertisements for the A & P last week? Yes No Do you remember Yes No (IF YES) read more than half read less than half seeing advertisements for Wrigley's last week? (IF YES) read more than half read less than half Do you remember seeing advertisements for Kroger last week? Yes No Do you remember Yes No (IF YES) read more than half read less than half seeing advertisements for National last week? (IF YES) read more than half read less than half Now, thinking back over the past three months, can you recall reading any newspaper advertisements for (EACH STORE)? A & P? Yes No Wrigley? Yes No Kroger? Yes No National? Yes No (IF YES) How often? about once a week about once a month How much usually read? (IF YES) How often? less than half more than half about once a week about once a month How much usually read? (IF YES) How often? less than half more than half about once a week about once a month How much usually read? less than half more than half (IF YES) How often? about once a week about once a month How much usually read? .___ less than half mnro khan ha“ ‘7 _7 . . I. ' . r! . . ." ‘i f '.I -' ‘, ., .' ° ‘ " n I t: , . .- \ - ‘ ‘r' . ' .‘~ . .. . I -»"V ’ . l ' . . I '. . .' 1 I I I I . . r , . I , . ., . V . t i I. Q . . . , - A , I ‘h-I. I . I ~ .-.. —.. . . . .e, ....~ .. . I' I a . . . . , . . ..I -... . l I , . I ‘ r - ,- - . .-- . I .. ’ . . : , ' ' . r .11 l _ - ‘ t ' . , \, “ e I'.‘ ' ' :-.- '. _ ' f... I'. -' I - ....A _ . ......-.. I. ' I (I . . I a ' .n c . l“ . « . , . I ‘ I , ' ' I , . ‘ -- ' 'SI . I I ' ' - r - , , . I a) I . ~ " I': 1 4‘ ,I . ( -. z ’I ,,..._ -I . ' j ' .sl 1 - . . _ ,. r I ‘ , _ ‘, . .. , .1. . «., (‘l‘ 'I L . 1 1 , A ‘a ' I . I .r . ) .‘ ‘ i . A . . .5 .. . v .' .‘ ‘ I , I‘ . . I z. . . .I - I . . x . ..... . I .. u . ’ . 1 ‘ 1&4. I II. - '3 ‘ .m ..- .. .. ~ . r . _ . .(’ I‘ ~ . ' ‘t . . I‘ .'-.‘ ‘ - " -‘ ' . .1 " a ~ ‘I -v . .'. - ' ~..; .I ‘- :. ‘I “it. w- 0“- . no .. m . v , . I If)... I '. l I ,. ,l. l..' .‘_.,, 1 c- I . . . ...v .. J...‘ ‘ . - I h. . .1 u ‘ ' r' - ‘ 'V I r I , . J :3 ' - I . ~ . I . . . ‘ I - ._. ...- .. ..__... A ' In I. )a - . . 3 . “ . ' - ' ‘- . , ’-. :.r o ' ' '1', , - .- a . . , . . _ , ' I. i-'."’ . . ‘i . Mm ~..~_., 'zf’u ‘ 3:. " I, N‘ D' ‘0. I .,- .. ‘I. ~ __ .' . L L; I! I . \ . .' v-I‘I“ \J ,. .- 1... .‘e‘ . . -. .- , . .. I l',~ .. .. v .I ' _ - . ." .. .7 .l . . I I (Iv-4‘ . . '-.1 . vv ‘v fl”... 1‘ . . . ' ' 4 I”. ' p ‘ ,’ 42 43 44 45 46 47 48 49 SO 51 52 $3 6.1 6.2 6.3 7.1 7.2 8.1 8.2 Page 5. (299) Do you have a radio? Yes No (IF YES) Do you remember hearing any radio commercials for (FOR EACH STORE, ASK SPECIFICALLY, ONE AT A TIME) last week? A & P store? Yes No Kroger? Yes No Wrigley's Yes No National? Yes No Now, thinking back over the past three months, can you recall hearing any commercials for (SPECIFY EACH STORE)? A & P store? Yes N0 Kroger? Yes No Wrigley's? Yes No National? Yes No Do you have a television set? Yes No (IF YES) Do you remember seeing any television commercials for (SPECIFY EACH STORE) last week? A & P? Yes No Kroger? Yes No Wrigley's? Yes No National? Yes No Now, thinking back over the past three months, can you recall seeing any commercials for (SPECIFY EACH STORE)? A & P? Yes No Kroger? Yes No Wrigley? Yes No National? Yes No Did you receive any super market advertising from (SPECIFY EACH STORE) in the mail last week? A & P? Yes No Kroger? Yes No Wrigley? Yes NO National? Yes No Now, thinking back over the past three months, can you recall roading any mail advertising from (SPECIFY EACH STORE)? A & P? Yes No Kroger? Yes No Wrigley? Yes No National? Yes No __, .—..- v .n .. . I . “‘ .‘1 v u.- . a——" O . ‘_._.- ll l2 l3 14 15 16 17 1.1 1.2 2.1 2.2 2.3 Page 6 (300) OCTOBER INTERVIEW: SUPER MARKET STATED FIRST CHOICE INTERVIEWER: IMPORTANT: THE STORE SHOWN ABOVE IS THE STORE RESPONDENT SAID WAS HER FIRST CHOICE STORE LAST OCTOBER. - CHECK TO SEE WHAT STORE IS STATED AS FIRST CHOICE NOW (page 2 question 1.2). '1— IF SAME STORE - GO ON TO PART II IF DIFFERENT STORE - ASK QUESTIONS BELOW. When we interviewed you in October, you stated that (see above) was your first choice. Now you state that (new first choice) is your first choice. Why did you stOp going to (OLD STORE)? (DID SOMETHING SPECIFIC HAPPEN--STORE REFUSED TO CASH CHECK, ETC.) Would you say, in general, that you were satisfied or dissatisfied with the (OLD STORE) at the time you decided to change? (CHECK ONE) ____very satisfied no particular feeling quite satisfied slightly dissatisfied slightly satisfied quite dissatisfied very dissatisfied llll How did you happen to decide to patronize (NEW STORE)? Before you made the change, what was there about (NEW STORE) that attracted you? Did anyone in your family say anything about (NEW STORE) before you changed? Yes No Do you remember what? Did you talk with someone other than a relative about the (NEW STORE) before you changed? Yes No What did the person(s) say about the (NEW STORE)? Would you describe the person(s)? (FIND OUT IF NEIGHBOR ON SAME FTOOR, OTHER NEIGHBOR; FRIEND, NOT NEIGHBOR: TEACHER: ETC.) ~-_-. 1 , . ---— n. 4 . ‘I ... u.- .— 1. .. ‘E . .,'v, _. H.,; 5,. 18 19 2O 21 22 23 24 25 2&5 217 213 2.4 2.5 2.6 2.7 3.1 Page 7 (301) Did you see any of their ads in a newSpaper before going there? Yes No (IF YES) In which newspaper? State News State Journal Can you recall if there was a particular ad which helped you decide to go to (NEW STORE)? Yes No (IF YES) What did the ad say? Did you hear their radio commercials before going to (NEW STORE)? Yes Nd (IF YES) What did you hear? Did you see their commercials on television before going to (NEW STORE)? Yes No (IF YES) What did you see? Did you receive any mail advertising from (NEW STORE) before going there? Yes No (IF YES) What did it say? Summing up, now, what was the most important source of information which helped you to decide upon the (NEW STORE), among the informa- tion sources listed below? (RANK FIRST, SECOND, THIRD, PUT "0" IF RESPONDENT SAYS DID NOT SEE ADV. OR DID NOT TALK) Advertising (in newSpaper, radio, tv, or mail) Conversations with relatives, neighbors, or friends Looking for myself Please rate the following sources of information. (PUT CORRESPOND- ING NUMBER IN BLANK IN FRONT OF EACH SOURCE) l - very important newspaper ads 2 - quite important radio commercials 3 - slightly important tv commercials 4 - not important direct mail spouse mother neighbor, same floor neighbor, other friend, not neighbor looking for myself .-§ ....... (302) PART II The last part of this questionnaire is designed to measure the meaning of different words by the use of the "rating scale" shown below. Study the example to see how the rating is done. Advertising healthy : : : : X : : : : sick By placing an "x" in one of the spaces you can show the direction and the intensity of your association with the word "advertising." For example, suppose you feel that "advertising" is neither "healthy" nor "sick". You would indicate your feeling by placing the "x" in the middle space as shown above. However, should you feel that "advertising" is very "healthy" you would mark the scale like this: Advertising healthy : X : : : : : : : sick If your feeling is that "advertising" is quite "sick" you would mark the scale as follows: Advertising healthy : .: =: : : :_§_:___: sick ‘ cfi _.* * However, if you feel that "advertising" is only slightly one way then you should check like this: Advertising healthy : : : x : : : :___: sick NOW, WOULD YOU MIND CHECKING YOUR.RATING FOR "MY IDEAL SUPERMARKET." lJ___; 'HY'IDEAL STORE good : : : : : : : : bad 18_____ ‘ fair : : : : : :__:___: unfair 19___ pleasant :__: : : :___:____:__: unpleasant 20__ clean :__:___:__:__:___:__:___: dirty 21___ friendly :_:___:____:___:__:____:___: unfriendly 22___ close :___:__:__:___:__:__:___: distant 23__ expensive :___:__:___:__:__:__:___: inexpensive 24__ progressive :_:___:___:__:__:___:__: conservative 25____ high class : : : :__:___:_____:___: low class ---NOW - TURN TO NEXT PAGE AND DO THE SAME FOR THE SUPER MARKETS LISTED. ---WORK AS QUICKLY AS YOU CAN. MOST PERSONS FINISH THE ENTIRE PAGE IN LESS THAN FIVE MINUTES. ---PLEASE CHECK EVERY ITEM. (IF YOU HAVE NO OPINION ABOJT A PARTICULAR SUPER MARKET LISTED, CHECK CENTER SPACE). . o — ‘ p . . ....-\I I' o ." o A, y. l .' ,l n” g . ..- ' "'7 JR I.,-‘ u w-- -—-t-o- ..---.o C o - w“. o a n .00" “Q I I . 7--.... I a u- l a . u': s n. . I . f I o n I .- I 4 l . I nun.-. .- ..p.n-. . ou-- - o. s . . o. . . . .--. ...—o. 00—. -- o- .4... . c ”an..- . ..-~— ‘ | .' » ,1 , a . .. n...- 'a . .- . . .. -.... fl...“ .4- g-.. - -.-uop_ 'l‘ 'l .| o n , . . ,. l . . 4 '. .c .n .,‘,4. . I , . l I l l u n l '1 (j. ,. In ' _. . . A . I . . . . . E. . .l. " .. 1 . ‘, , . ,... y . . . 3'!"")| a ;‘.’_ ' l ow-.- . o . . T‘: o . A v . a o .O. ‘I. ‘ or . . . .0 - . . n u -- p‘- u“- . . S., .. ?‘: ..- ....__.._ . a . . . . u '. . .._. .. I ....... ,. ..-- u . . . .. i .. .v. - I in .- .’. "r_'.-)-I .L I u-‘ .I ~ I . A d ‘l" KROGER A & P SCHMIDT'S NATIONAL WRIGLEY PRINCE BROS A & P SCHMIDT'S NATIONAL WRIGLEY PRINCE BROS KROGER SCHMIDT'S NATIONAL WRIGLEY PRINCE BROS KROGER A & P NATIONAL WRIGLEY PRINCE BROS KROGER A & P SCHMIDT'S WRIGLEY PRINCE BROS KROGER A & P SCHMIDT'S NATIONAL PRINCE BROS KROGER A & P SCHMIDT'S NATIONAL WRIGLEY KROGER A & P SCHMIDT'S NATIONAL WRIGLEY PRINCE BROS A & P SCHMIDT ' S NATIONAL WRIGLEY PRINCE BROS KROGER SCHMIDT ' 3 NATIONAL WRIGLEY PRINCE BROS KROGER A a. p good fair pleasant clean friendly close expensive progressive high class good fair pleasant clean friendly close expensive progressive high class good fair pleasant clean friendly close expensive progressive high class good fair pleasant clean friendly close expensive progressive high class fair pleasant friendly close progressive high class clean good expensive pleasant good close high class fair clean friendly expensive progressive O. .0 O. O. O. O. .. .i‘. a J O. O. O. O. O. O. } O. O. 06 DO 00 on ”U. .. O. O. O. O. I. O. "0‘ O. O. O. O. O. O. O. O. I. .0 .- J \ O. O. O. O. O. O. (303) bad unfair unpleasant dirty unfriendly distant inexpensive conservative low class bad unfair unpleasant dirty unfriendly distant inexpensive conservative low class bad unfair unpleasant dirty unfriendly distant inexpensive conservative low class bad unfair unpleasant dirty unfriendly distant inexpensive conservative low class unfair unpleasant unfriendly distant conservative low class difty bad inexpefi‘ls 1V9 unpleasafi»t bad distant low class unfair dirty unfriendly inexpensive conservative - — .. u u. ... . -u .- 30h APPENDIX D Reproduction of the Advertisements and Readership Scores from the Readership Penetration Test* State Journal Prince Brothers (Shop Rite) A & P Prince Brothers (Shop Rite) National A & P Kroger Wrigley Prince Brothers (Shop Rite) Schmidt's State News A & P Kroger *The nine ads appearing in the State Journal were used in the readership test for ads read "last week." The two ads from the State News are included to provide examples of the weekly ads carried in the college daily during the study period. 305 .1“ «a». lflgfli Noted Seen-AssoCiatedV Mos ea‘d ‘m‘flfi ‘347% 183% .112%.f g. In mm ’ Lam-0n 4:]P0RK 39 m ROAST a. 47m mom ...., OR SH UUUUUU rm (0(1er I u :3 ! La :2 ”$441129 . ‘ ;- - monilnous lfP:0n I; ‘/ m E:- ' 306 mountain: nun-nuns.“ “II "T0 m Social Security Chanchm-c to Affect Many hmum ddhflvmmb~o~oboolb*-.c~M'loo ~~dwbflw fifimmnmuuuwh MiG-Va“. W09? I! laid (I. hurt "M-fiirouh'b .,-',.w.wohv - “hunt—n-.. “In" M a.” M M '0 - .od '1... nth -~”“""’ ”"h‘ t d -- "- ~~ ' 04 IL! I..." ”fit (- Inn-no.4 lit-olcouo'wvm""~vi.m.~.-m ‘ ...'.”“ h... ‘ ‘- ”UV ‘O—t-h an on... can I. -- m Ibo-«Q. mod-u....b.“.m-w.w~|u.m—u~‘uaum ‘bOW‘J‘OM'~ -. . v-u-b'M'U'IUC'b--'~'-.~-d..mummw" hhu-U*M~M~~ .m u ‘w A”. “"hfi'oocq-b-fl-‘dn‘w‘NMDhW ' *- n-‘g; volt “up “a... in: V\. ,..'. ~,, ~H~mfil~Wndfl hIH.—b I—Mddodcdy- " ”m- ’d .I.' w- d O .“W-~‘Mrulv‘dbp-nu—nb-m‘u-m—m 'J‘f "0“' 0 " N"m~'.‘h t-hwvocdho u_h-—~H§-.O~GOD—' :fi~h~v- Ia. Mon-no.0... ~n---‘b In... n—ohh—lI-WW -0. .' IVs-u ”hmo-L "" "'""~‘ "m- 5‘. L...- -h who _' out. :- -._. L'csab-ocwu- "“'"~~.~-‘ *Iq-nm-nk-mc—n- 'V" 'u'. h -u' own M“. .10 gnu-1'0. 'Infifl. nqfin nhbho—ownnd-b. hp-q-mmdh \- "ud ~u-n HM J p.m- 'IlD-Ibolvud-I-‘u'u—w 0.0 I‘ N“. H‘ o-‘nh' h .01 t.-—‘\df--.."- MW . hfi-“¢.~lh~m you 0 m -o¢ DO. —.I .01—s1 91"... Ov‘N.I“" . .~ - b- .«u L.,. ' ‘ ' ' '“° ' .‘NI 0'": I‘d Io.- . h m W.,-d \-c‘¢A-. .—.- v . "‘°“'.~: .- \‘ 1 : v . .--.~d C-.|V.<.( \. 1. . , ' «O - N.-. ” m h n ., -b.. .r b ”*0 ~. ‘9‘1W‘ ' . “ -..\ a... . , . pfild¥snv - .‘ 'r . '5. .. .. Into a . ..... I‘ I D . p...- -h ~ [b I 4 ¢-‘o I.,. .-~‘" d I d I.,..- ' ' Ik."h3. '* 'I— ~- .l'n ’- l-rq . .c * to. U-. - -- '. .r.‘-l .— l.. .n w. .. VD: Disc-“u. .“‘. n!“ - .... .;.f. .. . ' sun nun .v n to" I- I SMOKED HAM - ‘ -_.,' m D ‘ “:11": . ' .J «In m Lawn u u we . . - . a m“ . . . ..: SPARE RIBS "thy "rail-II In “Wran- '\\’I.“" h" 3! O . Path: :2: '73. 79¢ Inlay (no... .312: 85¢ WWw1'c': 39¢ My“ am... 3.. 47¢ Id“ :2?! L05.” 37¢ w 3'3 .2: 25¢ “M “Wanner-av 23.2% cm Iim § mm ....... :4 ss. -—u-HuI—vb- m ”xxx. 0. ‘* 4...: ' may ":32: :3. 50¢ ‘ Ian's Mun- m a... 10¢ mu a... 1" h. Mu- .. ....... ~51: ‘2'" ..... 95mm- WM-‘a- antenna! loam: ”no. m- r "a: .3... "3'“ g; “a: .31... .2. .5. “SHOWBOAT” “55.27%” ~-----' 169 .. 833 .. Sal-numb. 32:. n. m“ hi“"°‘.."°‘32.%‘.2& "m , ‘ a»... “M ":37: ”sum mt-xzwu. '2'.“ ..— ‘ "“'. :03; ;; J... 4237. a" doxolo Oil 2 hint-lulu! 13- W” or.-. (so; ‘5‘ 2 ‘43.“ m 0:.- II. 59‘ .' ..I- ' ' '.-..'..-.‘h. - ---------------.... col '00... d .Hvflo '0va “a------— --- O'LWIE uw-muwrhv... ”I. .LI.’ wtuntu- “a III. I I finitflm Y-OH. ' .LI 0... “Albmfi‘ uhhgno~nzonn 307 unfit-tubby”. . *VIIP wfi, ,__ . w. Noted Seen-Assoc1ated Most Read ‘ OUR OWN HOME ”05- BULK LSAUSAGE "’ 29¢ PORK STEAKS --- 492 . PORK « fl ROAST "‘ “’1 Whole PORK Shoulders " , SLAB BACON "‘ 392W VALUAILI ACE-HI OI" 8TH” WITH EVERY FOOO rum 308 -mnuu—uhlm ' Noted Seen-Assoc1agad Most Read 408% ...29.5%f‘{ ' . . o HOUSEK 1-. ammo camps 19:5,.“ NAnouAi "imifimm an.» IB” P433: 98: sumo": 98¢ ’(o—p—‘HH 0'0 ‘4-l I , mus “-.:",,.,.."'":...... Em“; ‘10; m5":”" mus "in" 7.0”! tic mum-Taw— mun“ ‘ ------- .mnm c-\u\o ”“""“"'"“ If!!!” Ital-a0... 11:. “Ht“?! g... WW‘APPLESMA 2 Wm- ............. "'"."Eo'o'iu' "WW "} FLOUR ...:= 3» 19:.”3,=-3159 ”:‘tz‘: 333nm ____________________________________________________ 1 ................ mm?“ “1:“ ”mamas , uptuukscq g; WM gmfjfc '3'; “4365..., ...... 1-12:3” '4‘ 33‘ @E W “mu - “um“. mu 2,: ,' roam TISSUE . 'm -.:- n. "‘ ‘ 4 u. 39' 5‘6“?“ :1. a: If"... ------------------- — ‘1: u an anus NI 5 T M a: "I mu" m : MAR'GARINE I: 3 o... m It“ ‘ |=- 31¢ :L 2h“ 39' : ~00 --------- -u-g----o--------J xmoxu. 00-0p Be“ ”Wu ”(i . ‘20...“- mm 50 Aw our mt: CHIP ’N DIP _ 25, 11x ,7. i; ""855 -._.. ..._...... ”1.29:“ ._..“}...... 0 $29.“. u '7: man $7.30AP “SF-‘0; RT;AIW MY SOAP "Inn SOAP m m '2: 33: ‘z: a: = 23: “a: 1: 19¢ 1... '- 35: = 43: :- =49et :33: “4'3“ “.233: 2:49: 2:21: 2:31: a: mun-aunts; m.- WHIP” a BIS to Get ‘A' Slay er, 16] “WON“!- albmlfl .bl. ll\l'b'fll|l . "n i. I . ., o‘ h. m vc") A u u. or... “on ~II¢~ — _ Divorce Suit 0“ I :AJf‘ :gi O :3 3:3; uI9¢ gains... a 89¢ kAD cw , nun-on- “. STEAK SALE! 5Q-0%, . 309 - -.:n _,' 1 1:0. f IIOLE. cOIPLETEIJ clEAIEI ‘vv :3? — WW v- k Noted Seen-Assoc1éted Most Read h0.7% " 27s g,‘ SMOKED HAMS 7 29 05% " A .- . :3. at I 5: a 31' ' 3‘89: .., .: “.. u. 39‘ ll. 7.6‘ wwN—I’L’" hilt “dive-Midi “5.334" .... 5 °' 'w uI-fi o. -:& as; a L 'w unnum- ~EféépoTATOES A A I I . , ‘ . A '0 I . A. .o‘ l . s \ In: I...“ 0'. I . I. a 1' 0.1 ‘ " u . l .gv.. . D 'O u n t . “ '- -¢... . -~ I . s.v\-I n. - I l I . . \.. . .. . >1 \ I. A u O '90 I. o ‘ . - .v o ‘4'“ ’ O to ' I h. A ”Imam II\I¢\I\‘T mm II. M M $372 29¢ (.1! MM “'77 "' 39¢ (1,1272 29¢ AIIII nu mvmowflmmm-Iooo 4:“ 99¢ Mudw MUM-nu '20. IBM" sum. Ira“ .3. W.m‘:6&“.79¢ lady“ nonu- “42¢ hulfiwwI'glh Pith“ can. 32* am—wm-r ~h‘oquI IDA “n " W' 7 0' "3 :- OHOWIOI' ONIONS “‘91“wa 47k a..." c “'. . .69 ll. 10‘39 GIIPGII'IIII Jlllce Sal-(Ilia m a: a: India-II :1. R33 79‘ haul-II. nun- “: 55¢ '- an“ 1“,“ hhanu-u‘” “*1‘:L‘:‘m ‘0 2(‘3 2" M "émummuda’a’, Hutu-aw...» "I r!“ I123.) 3 5.”? 29¢ lac-kl" mum 0.19“ ._._.—-,, ‘_ b- r 1"" ' baht-.m- ”mm “‘ . “an hard-w Meagan—a On“ I." "In“ CHERRY PIE "'"""""' "'"' H“? ‘ In” “me mun-u .. 39¢ W. CO-Op Read I“ u If": {1 5.9 "W“ III-n . III Li?“ Ind ":1. ‘53.“: 2 ...... as: u H. m ‘. 12.2% A I“ 57‘ ,- - a “ :za“: innu- _-- sum—haw :0 ”W m ‘::'::r 3““ 9L" ‘3‘“ a: , a 7" u: 33: ’3: ‘79 ”'4: 7 t h ummnmum-J: I: III 2': ‘I '2: “I :. 9¢ In” —_ ”wuflOh-DOM *Ot‘bf'.‘ ~HHdb~OAIbOII “MdhOAI-I'.‘ .12.... 8.8.3 .SIIu 3.34.4.» 34.3.1533 9.9-Illa} . l!’ I... z. 3...]... I 38.8.3 .3 1. 1.3.3 .35... 1.31.3.8 _ in... 1.9.3 8.... .3. 11:11:... II“ «1:06. it] I.|I .1190 13.. 8r! 5...... 3n .1 .34.. U 1: :2. 1.2.15: " «an ... 2.9.8 n3 1:... 1.3.18 2... H. II. . II . I. II I I . y , . u-..,a”m_”..._..m \: .I ZOO...OIO..EM .I.I.n.. . .51 x .. .. . . . .......... . 1‘ . . . H. .os . .211. . . .. . .u u .a. . 1... .- .l; q . o. ..l 13.21.31: 00. 9»: ....\.ll1l. \\ -- 1 3n .. .; 3.....ahow. 3. . :5 SHMII x: . £3 .216. .1: 11...... 5.1.43 1 \I 332......136 .3 . .23.” am a. .3... 2. .. .3.» a. .3. £13.... a... Sam I a! .(3 Spark- >w§ 1!» (.5503 I W I'M . r .m h... IH... 71...... r. I $5.132 .I.I.I. .1... 1 11...... .1... h 1...... 11.11... _ p . 11.11 . 1...... {In .III((I¢\II.IIII\ I/(II(\ . I‘M/III M . . u. . 7. :1.-. WWII. .im.cH.JEOofod, P. “A..."wnim—PI? '1‘ IInvII.1rI::.h vmmx pmoz Umumfloomm