CONSUMER ATTITUDES TOWARD WINE

CONSUMER ATTITUDES TOWARD WINE H. G. SCHUTZ and J. H. ORTEGA Respectively Professor of Consumer Science and former graduate student, Department of Consumer Sciences, University of California, Davis, California 95616. The financial support of the Wine Advisory Board and the advice of their ad hoc Committee on Consumer Research is gratefully acknowledged. Accepted for publication January 25, 1974. ABSTRACT Fifty-six beverages (43 wines) were rated against 48 use-attributes by 52 Woodland (California) housewives. Mean appropriateness and percent appropriateness values were compiled for each beverage over all use-attributes. Factor analyses were conducted on wines and on use-attributes. Three wine factors were differentiated on the basis of occasion served, how used, and psychological qualities. The three use-attribute factors were selforientation, other-orientation, and functional orienration. INTRODUCTION The recent rise in wine popularity (2,3,8,13) has led to a concurrent rise in time and money expended on communications concerning wine. The way in which communications are received by the consumer is largely determined by attitudes. Therefore, information about consumer attitudes should prove useful in formulating more effective communication so as to fit existing products more specifically to intended use, to develop new products to fit unfilled uses, and to locate areas where consumer education may be needed. The specific goal of the research was to show how the consumer perceives wines in relation to their uses or attributes. Additional impetus for an attitude study was provided by the lack of such studies in recent literature. The most recent attitude surveys discovered were in 1955 and 1967 (1,7,9,10,11). It is hoped that this study will widen understanding in this field. MATERIALS AND METHODS Sampling method: A 1967 census tract map of Woodland, California, was divided into sections on the basis of median annual family income: $5,000-7,000; $8,000-10,000; and $11,000 and over. Section numbers were drawn randomly, with an equal number drawn from each income level. Two addresses from each section were chosen randomly from a street address map. Surrounding addresses, as well 33 as those chosen, were written down. Names and telephone numbers from the selected addresses were collected from the city directory. The final list was composed of housewives only, since it was decided that they represented both a relevant and accessible consumer group. Each respondent received a letter, the type depending on whether it was for the first or second phase of the study. Several days later the respondent was telephoned and asked if she was willing to participate and was between 20 and 50 years of age and a member of a family which served wine two or more times a month. These criteria limited the sample to families that were regular wine users. First phase" The purpose of this phase was to collect consumer's ideas on wines and the useattributes of wines. Thus, an open-ended questionnaire format was used. The respondent was asked first to name all wine types she could think of. She was then asked to suggest what she would use each type for, and lastly, what other product she might substitute for each use. Of 162 respondents telephoned, 23 completed the questionnaire. Second phase: The questionnaire for the second phase was derived from the analysis of answers to the first questionnaire, plus suggestions of relevant items (beverage) and/or use-attributes from an act hoc Consumer Research Committee of the Wine Advisory Board. The items were selected as representative of the following categories: wine (table, dessert, other) ; nonalcoholic beverages; and

WINE ATTITUDES--34 alcoholic beverages other than wine. The total number of items selected was 56 (Table 1). Use-attributes also selected as representative of general categories were time of day, occasion served, where served, physiological factors, how used, psychological factors, persons served, and physical characteristics of products. The total number of useattributes chosen was 48 (Table 2). Combinations of items and use-attributes were rated on the questionnaire by using a seven-point appropriateness scale anchored on both ends. If the respondent was not familiar with the item, she was asked to encircle the item number and rate it according to what she thought it was like on the basis of its name. In addition to the 56 items and 48 use-attributes, the questionnaire included a page of instructions and a page of background questions. The questionnaire was delivered to each respondent and picked up at the end of one week. RESULTS parison). Whiskey's perceived uses were "for cocktail hour," "when entertaining," and as "a stimulating drink." Beer nearly tied with whiskey (92% versus 90 % appropriate) for use "by men"; in other respects it was neither high nor low in any caregory. Wine name --- general to specific: Two sets of wine types were compared to determine the effect of specificity of name. The red table wines chosen were a red wine, Burgundy, and Pinot noir. The white table wines selected were a white wine, Chablis, and Chenin blanc. In the red category degree of appropriateness declined from general wine names to specific, though the patterns were similar. Chablis had a higher appropriateness rating than TABLE 1 List of Questionnaire Items Mean appropriateness ratings and percent appropriate values (percent of people rating five or higher on the apropriateness scale) for each itemuse combination were calculated and grouped by use categories in a series of tables. These tables are too extensive (2,688 values) to be presented here, but are available from the senior author upon request. Some general patterns evident in the data are summarized below. Wine and other alcoholic items use appropriateness: Three wine items (a sweet wine, a table wine, extra-dry Champagne) and two other alcoholic beverages (beer and whiskey) were compared on the following representative uses: time of day ~ "goes well with meals," "for cocktail hour," "something that's good anytime"; occasion served m "to serve when entertaining," "for special holidays"; where served --"to use at home," "when I'm at a restaurant"; physiological ~ "something that's good for me," "a stimulating drink"; how used "hard to to serve correctly"; psychological- "to drink when celebrating," "when I want to feel creative," "to impress someone"; person served ~ "for men," "for women," "for young adults," "for the elderly." Table wine as a general item was perceived as appropriate 'with meals," "when entertaining," "at home," "for men and women." It was not seen as particularly healthgiving or stimulating. Sweet wine was rated as highly appropriate for only a few uses. Sweet wines are discussed later in comparison with other wines. Extra-dry Champagne was rated highest in appropriateness among the wines "for cocktail hour," although, even then, only 68% viewed it as appropriate. In comparison with all other items Champagne was perceived as second only to whiskey in its stimulating properties (this is, of course, limited to the five items chosen for com- Whiskey A red wine Burgundy A green bottle of wine Rhine wine Chablis Sauterne "Zinfandel' Fruit juice A white wine Cold duck Sherry Extra-dry champagne Gin A half-gallon bottle of wine Apple wine Imported wine 'Riesling' 'Pinot noir' A clear bottle of wine Vodka 'Gamay' 'Green hungarian' Port A magnum bottle of wine 'Sauvignon blanc' 'Grenache' ros6 Fruit-flavored wine A d ry wine Soft drink (soda pop) Scotch S6millon' A sweet wine A half-bottle of wine (tenth) Slightly sparkling wine 'Cabernet' Punch Wine in a Burgundy bottle Coffee 'Chenin blanc' A fifth-bottle of wine A dessert wine Cocktails 'Gamay' rose A brown bottle of wine A table wine 'Chardonnay' Liqueur Tea A ros~ wine 'Pinot blanc' Bourbon A gallon-bottle of wine An aperitif wine Brut champagne Beer

35--WINE ATTITUDES TABLE 2 Comparison of Major Wine Types on all Uses-Attributes Degree of appropriateness Uses-attributes Red White Ros~ Sparkling Sweet Time of day goes well with meals for cocktail hour to drink before bed something that's good anytime as an evening snack good with lunch Occasion served to serve when entertaining for special holidays at parties Where served to use at home when I'm at a restaurant Physiological as an aid to indigestion to help me sleep something good for me when I'm watching my weight a stimulating drink picks up my appetite when I'm not feeling well How used to drink straight mixes well in cocktails as an appetizer (aperitif) goes well with fish or poultry an after dinner drink good over ice goes well with meat to cook with hard to serve correctly Psychological when I'm sad when I want something different to drink when celebrating something I really like when I want to feel creative promotes sociability a relaxing drink to impress someone gives feeling of well being Person served for men for the elderly for guests to drink at any age for young adults for women Physical a refreshing drink as part of daily meals a drink that is soft a fragrant drink easily stored a light drink high high high mod low low mod low high low mod mod. mod mod low high mod mod high mod mod low mod low low low low low mod low low high rood mod low low low low low mod high mod high mod low mod low low low mod mod mod mod high low mod mod mod high mod mod mod mod high low mod mod mod high mod low low low mod low mod mod mod high mod mod mod mod low low

a white wine, but Chenin blanc was lower than a white wine. The patterns of use were, as above, similar for all three. Comparison of wine types on all attributes: The comparison in this case was for percent appropriateness values on basic types of wine. The types selected were: red, white, ros~, sparkling, sweet. The values were divided into high, moderate, and low, using cut-off points of percent appropriateness: high, 100% to 61% ; moderate, 60% to 31% ; and low, 30% to 0% appropriateness. These results are shown in table 2. The use attributes are grouped by the categories given eariler in table 2. Time of day: All wines were rated low on "to drink before bed" and "as an evening snack." Red, white, and ros~ ranked highest on "goes well with meals," with sparkling rated moderate. Sparkling was the only one rated high "for cocktail hour." Occasion served: There were no low ratings in this category. Red and sparkling rated highest over all attributes. Where served: As in the above category, ratings were uniformly high. The only rating of moderate was for sweet wine "when at a restaurant." Physiological factors" Low ratings were given to all wines for these uses: "to help me sleep," "when I'm watchng my weight," and "when I'm not feeling well." There were no ratings of high, and in other categories moderate ratings predominate. Sweet was rated low for all uses except as a stimulating drink. How used: In general, all the wines co-vary over these uses with the exception of "goes well with fish or poultry," "goes well with meat," and "as an after-dinner drink." With fish or poultry, white is high, and ros~ and sparkling moderate. With meat, red and ros~ are high, and white and sparkling moderate. 'As an after-dinner drink," sweet is rated moderate and the others are rated low. Psychological factors: Red, white, and ros~ followed the same patterns of ratings over all of these attributes. Low ratings were given on "when I'm sad," "when I want to feel creative," and "to impress someone"; the remainder were rated moderate. Sparkling was rated high on "when I want something different," "to drink when celebrating," "something I really like," and "promote sociability"; low on "when I'm sad" and "when I want to feel creative." "When I'm sad," "when I want something different," "something I really like," "when I want to feel creative," and "to impress someone" were rated low for sweet; the rest were rated moderate on these attributes. Person served: All wines over these categories were rated similarly: for men, high; for the elderly, moderate; for guests, high; to drink at any age, low; for young adults, moderate; for women, high. WINE ATTITUDESm36 Physical" "A drink that is soft" was rated as low for red, white, and ros~ wines, and "easily stored" was rated high for all three, with the other caregories all rated as moderate. Sparkling was rated high as a 'refreshing drink"; the others got moderate ratings. General results: Viewing the results as a whole, some trends emerged. Wine items were rated as 90% appropriate or above less frequently than nonwine items (21% of wine items vs. 77% nonwine items). This trend is based on the fact that seven of the uses were rated as 90% appropriateness or above on wine items as compared with 16 nonwine items. Furthermore, uses for which wine items were highly appropriate tended to be similar to one another. They are: "goes well with meals," "goes well with meat," "to use at home," "for special holidays," "to drink when celebrating." Factor analyses: One way of finding meaningful groupings of variables such as the beverages in this study is to conduct a statistical analysis such as factor analysis (12), which determines the basic dimensions within a set of variables. In other words, a cognitive grouping of beverages based on the relationship among variables over the use-attributes in the study. To obtain a matrix suitable as input for the factor analysis, a measure of similarity, D, was computed between variables across all uses for the respondents (12). These D values were divided by the maximum possible D for the scale and subtracted from 1 to produce new values ranging from 0 to 1, with 1 being identity. These new values were entered as correlation coefficients in the factor analysis program (4). Item groupings- First, a factor analysis was run on all items including nonwine items. This yielded three factors identified as: 1) table wines; 2) hard liquor; and 3) sweet wines. Since the major purpose was to discover groupings among wines, a second analysis was run. This included only wine items and those use-attributes which discriminated among wines. Discrimination was defined as a range of two points or more from the highest to the lowest mean appropriate rating of wine items. Four factors resulted from the second analysis, two strong factors and two relatively weaker ones. In general, they were differentiated on the basis of these attributes: occasion served, how used, and psychological qualities. This conclusion was reached by comparing mean appropriateness ratings on attributes for items loaded highly on each factor. Factor one was designated "popular with meals." Table wines have high factor loadings, and both champagnes and sweet wines have low loadings. Factor two represents wines that are popular when a meal is not served. These wines were considered appropriate for informal entertaining. Port, a sweet

37mWINE ATTITUDES wine, apple wine, and a fruit-flavored wine were loaded highest. Low loadings were given to Riesling, Burgundy, and cold duck. Champagne loaded highest on the third factor, which was called "special occasion." Sherry ranked second-highest on this factor. Factor four is the weakest and was named "high usage inexpensive." Highest on this factor is Burgundy. Pinot noir, Sauvignon blanc, and cold duck are among those with low loading. The cumulative proportion of total variance accounted for by these four factors was 94%. Use groupings: Although perhaps less directly applicable to the purposes of this survey, it was nonetheless felt that patterns of attributes would prove helpful in tailoring wines to uses that are harmonious, i.e., belong to the same factor. The same method of factor analysis was utilized as previously discussed and yielded three factors. Factor one is "self-oriented." It is a strong facfor, with the highest loading being "when I'm sad." This factor involves uses that imply changes desired in either a psychological or physiological condition of the consumer. Other uses with high loadings are "when I want to feel creative," "when I'm not feeling well," and "to drink before bed." Factor two, "other-oriented," indicates uses that involve people in sociable groups and on special occasions. The high loadings on this factor "when entertaining," "for guests," "to use at home," demonstrate that this is also a strong factor. Loaded low on this factor are "when I'm sad," "when I'm watching my weight," and "when I'm not feeling well." Functional orientation characterizes factor three. These uses are part of every-day routine and not distinguished by size and composition of group, or by mood of the individual. The high loadings on this factor are "goes well with meals," "as part of daily meals," and "goes well with meat." Although weaker than the other two factors, number three accounts for 2% of the total variation, and all of the factors together accounted for 85'~. DISCUSSION Wine is perceived as appropriate primarily with meals but according to rules of "correct" usage, i.e., red wine with meat, white wine with chicken or fowl, etc. The range of use is widened by the utilization of wine on special occasions and when entertaining. These two points together indicate a knowledge of wine usage in the traditional sense or in situations where the suitability of wine would not be in question. Another asset is the characterization of wine as a nonstimulating drink. This could provide reassurance to consumers who are wary of alcohol, and lend appeal as a moderate middle-of-the-road drink. Since wine was described as light, fragrant, and refreshing, this might reinforce the above impression and provide a taste advantage over other alcoholic beverages. Another significant asset is the perception of wine as appropriate, at least moderately so, for all ages. Because wine is thought of in the above-described traditional sense, it is not used in other situations, in more innovative ways, or on a routine nonspecial-occasion basis. Although "popular between meals" and "special occasion" were factor groups, it was felt that, in view of the other results, wines in these groups would not be selected when beverages other than wine were available. Wine did not appear to have high prestige value for this group of consumers. This is a drawback, when taken in conjunction with the uncertainty of consumers about specific varietal wine names. Since varietal names and production place commonly affect price-prestige value, perhaps lack of knowledge in one area has led to a faulty image in the other. Another point on the debit side is that wine is not perceived as particularly good for a person (45% white wine, 42% red wine). Referral to the "self-oriented" uses of factor one shows that the uses with highest loadings were not deemed as appropriate for wine. This suggests an area of uses that is as yet unfilled. The image of wine use that becomes discernible is one of perception of wine as appropriate for a limited scope of uses. These uses seem to be largely determined by past patterns of behavior. The assets of wine, logically, should be emphasized, and the liabilities minimized. In doing this, there would exist the strongest possibility of wine image change. This is not to suggest that wine can be made to seem all things to all people. However, by the promotion of different wines with different appeals or emphasis or assets, a gradual change in image could take place. This change might be achieved through advertising tailored to product use, continued consumer education by such groups as the Wine Advisory Board, and further research into wine attitudes of the consumer to assess change and suggest further directions of research. The results of this study, naturally, can apply with a high degree of confidence only to the housewives who participated and the wines evaluated. However, it appears reasonable to assume that, considering the manner in which the sample of housewives was selected and the broad range of wines included, the results could be generalized to other housewives and wine products. LITERATURE CITED 1. Amerine, M. A. and C. S. Ough. Sweetness preference in ros6 wines. Am. J. Enol. Viticult. 18:121-5 (1967). 2. Clark, Richard H. The case for marketing 'super premium' wines. Wines and Vines. 52(2)9:43-4. (1971). 3. Clifford, James O. Wine sales go pop. Sacramento Union E8, February 19 (1973).

WINE ATTITUDES--38 4. Dixon, W. J., ed. BMD Biomedical Computer Programs. Berkeley: University of California. 5. Elmo Roper & Association. A study of people's attitudes toward and usage of wine. Sponsored by the WAB, August (1955). 6. Filipello, F. (Project director). Wine Consumer Analysis, Cummings and Assoc., sponsored by the WAB (1957). 7. Filipeilo, F. and H. W. Berg. The present status of consumer tests on wine. Am. J. Enol. Viticult. 10:8-12 (1959). 8. Folwell, Raymond J. and Richard T. Dailey. U.S. table wine consumption, state regulations, and market potentials. Amer. J. Enol. Viticult. 22:210-14 (1971). 9. Nelson, K. E., G. A. Baker, A. J. Winkler, M. A. Amerine, H. B. Richardson, and Frances R. Jones. Chemical and sensory variability in table grapes. Hilgardia 34:1 42 (1963). 10. Nelson, K. E. and H. B. Richardson. Relationship between flavor and maturity in consumer acceptance of Thompson seedless. Blue Anchor 44:43-5 (1967). 11. Ough, C. S. and G. A. Baker. Small panel sensory evaluations of wines by scoring. Hilgardia 30:19 May (1961). 12. Rummel, R. J. Applied factor analysis. 617 p. Northwestern University Press, Evanston, Illinois (1970). 13. Winship, Frederick N. Look for wet revolution at party time. Sacramento Union, E8, February 18 (1973). Amer. J. Enol. Viticult., Voi. 25, No. 1, 1974