Food Label Information: What Consumers Say They Use and What They Actually Use

Robert J. Vandenberg, Food and Drug Administration and The University of Georgia
ABSTRACT - The purpose of this study was to assess the following question: Do consumers' self-reports of food label use correspond with what they actually use? Evidence provided here suggests they do. However, the extent of usage of specific information components of the food label (e.g., the ingredients list, the nutrition label, etc.) is dependant on individual differences.
[ to cite ]:
Robert J. Vandenberg (1981) ,"Food Label Information: What Consumers Say They Use and What They Actually Use", in NA - Advances in Consumer Research Volume 08, eds. Kent B. Monroe, Ann Abor, MI : Association for Consumer Research, Pages: 484-487.

Advances in Consumer Research Volume 8, 1981      Pages 484-487

FOOD LABEL INFORMATION: WHAT CONSUMERS SAY THEY USE AND WHAT THEY ACTUALLY USE

Robert J. Vandenberg, Food and Drug Administration and The University of Georgia

ABSTRACT -

The purpose of this study was to assess the following question: Do consumers' self-reports of food label use correspond with what they actually use? Evidence provided here suggests they do. However, the extent of usage of specific information components of the food label (e.g., the ingredients list, the nutrition label, etc.) is dependant on individual differences.

INTRODUCTION

In 1978, a nationally representative sample of 1374 respondents was interviewed in a survey sponsored by the Division of Consumer Studies within the Food and Drug Administration. Although several other food related topics of consumer interest were addressed, the primary purpose of the survey was to assess consumer interest in, use of and concerns with various aspects of food labels, including the ingredient list, nutrition information, package claims, open dating and other related components.

The results revealed that 76% of the respondents claim to pay some attention to the ingredient list of food products (Heimbach 1980). Additionally, over 70% (54% of all interviewed) of these consumers reported using ingredient list to avoid purchasing certain food products. In contrast to those reporting no attention paid to the ingredient list, attendors were more likely to be in good health, be very well nourished, be white, be between the ages of 31 to 59 years, be college educated and have above average incomes. The majority (94%) of the respondents also reported using open date information to tell them about the age or freshness of the product (Hackleman and Heimbach 1980). Furthermore, 80% of this group (74% of all respondents) stated they use date information in making purchase decisions. The users were again more likely to be in good health, be white, be between the ages of 31 to 45 years, be college educated, and possess above average incomes. Finally, 64% of the respondents stated they paid attention to the information on the nutrition label (Heimbach and Stokes 1979). Again, these consumers were characteristically more educated, between the ages of 31 to 59 years, white, and had higher incomes (Baumgardner, Heimbach and Stokes 1980).

Although these results are revealing, they do not address some fundamental questions concerning consumer food label use behavior. Basically, there was no means provided through which consumers' self-reports of food label use and actual usage of food labels could be independently assessed. In other words, is there a difference in what consumers say they use and what they actually use? One possible answer could be that respondents were answering the questions in a socially desirable manner (Cozby 1977). However, consumers who report attending to and using the food label are generally more educated and possess higher incomes, indeed, of consumer demographics, education and income [Education was the best single predictor with income being second best.] account for the majority of the explained variance in reported use and satisfaction with food label information (Baumgardner, et. al. 1980). Aside from education, another factor could be the consumer's degree of concern with food label information. Some respondents may simply be more concerned with the quality of the food products they purchase and, as a result, they will be more likely to use the components of the food label.

An extensive research project was designed, therefore, to further explore some of these issues. The basic design (to be detailed in the following section) was a two (high versus low education) by two (high versus low concern) factorial. A subsample (100) of the original nationwide sample of participants was accompanied on a shopping trip by an interviewer. The major purpose of the project was to pinpoint exactly what types of products were purchased, why the consumer made the purchase, and most importantly what information was used by them to make (or not make) a particular purchase.

METHODOLOGY

Sample

Respondents for this project were selected on the basis of three criteria. The first criterion was whether or not the respondent agreed during the original survey to participate in one or two follow-up interviews. Out of the original 1374 interviewees, 557 revealed a willingness to participate again. The second criterion was the respondent's education level (one of the independent variables). Respondents were equally divided into those relatively well educated and those less educated. Operationally, the well educated group possessed one or more years of college, and the less well educated had never attended college. The purpose of this classification was to include both respondents who are perhaps more articulate and more readily able to utilize package information and those who may have more difficulty using the information.

The final criterion (and second independent variable) was the consumer's concern regarding food label information. This was operationally defined on the basis of their responses to 11 questions from the original survey (Table 1).

TABLE 1

QUESTIONS USED TO SELECT THE HIGH AND LOW CONCERN GROUPS

The respondent's answer to a particular question was scored, and all scores were summed to provide an overall Concern Score. Using this score, participants were then assigned to either a high concern category or low concern category. It was originally intended to acquire 100 participants with a distribution as shown in Table 2. However, due to faulty interviewing, equipment failure, etc. the final sample included 77 respondents distributed as shown on Table 2.

TABLE 2

DISTRIBUTION OF DESIRED AND FINAL SUBSAMPLE OF PARTICIPANTS

Interviews

The overall interview procedure consisted of three components. The first component was an at-home interview during which the participant was asked a number of questions concerning meal planning, health problems, uses and concerns with food labels, planning for the shopping trip, and avoidance or preferences of certain food types. This portion of the study was designed to provide a profile of the consumer's shopping behavior, meal planning, etc. before leaving for the supermarket.

The second component of primary interest involved the interviewer actually accompanying the respondent on a regular shopping trip.

The interview was completely open-ended and the interviewer both tape recorded the ensuing interactions and took notes on relevant behaviors. Shoppers were asked to talk their way through the food purchases as they occurred in the store and explain in detail how they shop and use information. Shoppers were asked to describe and explain any and all sources of information used to make purchase decisions. This included information the shopper has learned from past experience, recent information from source outside the store, (e.g., newspaper ads, information from shelf, tables or other locations within the store or information on the food label itself). The observer would prompt the shopper from time to time to clarify any confusing information.

The third component of the interview was again conducted at the respondent's home. First, the respondents were probed to determine whether they had modified their shopping behavior in any way, given the nature of the previous task (being accompanied on a shopping trip), Second, the respondent was asked (without seeing the product) to recall any label information on some of the products just purchased. The respondent was also asked to state any problems he/she might have had with the food label, anything that was of particular use and the interview terminated with a series of questions concerning food and meal planning. The taped interviews were transcribed into small booklets for ease of data coding so that type of product, information used for the product, whether or not it was purchased, etc. could be represented numerically.

RESULTS

It should be noted at the outset that analysis of the data obtained in this study has barely begun, and what follows is a progress report including some preliminary summary data rather than a completed research report. Much more extensive and detailed analysis remains to be done with the database.

Descriptive data

Table 3 depicts a simple frequency count of the top 19 products purchased by respondents. The 77 participants purchased a total of 1432 products with a mean of 18.6 products per shopper.

TABLE 3

THE TOP 19 PRODUCTS PURCHASED ACROSS ALL RESPONDENTS

As seen on Table 4, consumers used a total of 1132 pieces of information to purchase the products, yielding a mean of 14.7 pieces of information per shopper or about 0.8 pieces of information par product. This total does not include two types of information which were assumed to be virtually 100% used and therefore not coded: product identification (milk, cereal, etc.) and brand name.

The top category of information used (35%) was "Other", including such classes as label claims, grades, product appearance, product description, etc. Of more importance though, was the weight placed by consumers on price in comparison to the ingredient lists and nutrition labels. Thirty-four percent of all the information was price relevant, with ingredients and nutrition information accounting for 12% and 4% respectively, Furthermore, open date information accounted for only 4% of the total information used by consumers. Finally, 11% of the information used by consumers was size and quantity of the product. This latter fact was probably related to price since consumers generally compare the size and quantity of products to set a better bargain. Again, of course, this is a preliminary analysis: not all information is available on all food products. Open dates, for example, are commonly not available on packaged foods.

TABLE 4

INFORMATION CLASSES

Inferential Data

Analyses of variance were conducted on four components of the food label: (a) the ingredient list; (b) nutrition label; (c) the open date; and (d) the price. These components acted as the dependent measures in the analyses. Scores were computed such that one point wen assigned per unit of information. For example, a shopper using the ingredient list 5 times for 5 different products during the trip was given a 5 for that particular dependent measure. The same process was used for the remaining three variables. The independent variables for the analyses were the two factors, education (high and lay) and concern (high and low).

As seen in Table 5, only the results obtained from ingredient list use were significant to a large degree. The main effect for education was marginally significant [F (1,73)= 3.26, p<.07]. Consumers with a higher level of education used norm ingredients infatuation (x = 2.33) than those shoppers with a lower education level (x = .96). Greater significance was obtained then the concern level of the participants was analyzed [F (1,73) = 7.33, p< .01]. Consumers who reported a greater concern for more information used significantly more ingredients information (x = 2.67) than those displaying less concern (x = .62). The effects of education and concern level on ingredient list use was enhanced even more when the two factors interacted [F (1,73) - 19.93, p<.O01]. This results because education had no real effect on ingredient list use when consumers were not concerned with it. However, participants with a high level of education and a high concern for more information increased their use of the ingredients list substantially. Furthermore, even the less educated used the ingredient list more when they reported greater concern.

TABLE 5

RESULTS OF THE FOUR TWO-WAY ANALYSES OF VARIANCE

Although ingredient list use was affected by consumer education and concern, nutrition list, open dating and price were not particularly influenced by these characteristics (See Table 5). Concern level had a marginal effect on nutrition label use [F (1,73) = 3.11, p<.10]. Again, those who displayed a higher concern were more likely to use the nutrition list (x = .92) than those with less concern (x = .24). On the other hand, education had a marginal effect on open date use [F (1,73) = 2.92, p<.10]. More educated people were likely to use open dates (x = .83) than those less educated (x = .37). Education and concern levels had no effects upon participants' use of price information.

DISCUSSION

These results tend to parallel those obtained in the past and suggest that self-reports of overall food label use correspond to a large degree with actual usage of the information. However, use and understanding of certain components (e.g., the ingredients list), were mediated by individual differences.

First, in a 1978 FDA Food Labeling Survey consumers were asked to rate on an 11-point scale how much they attended to various attributes of food products (Hackleman and Vandenberg 1980). These attributes included such aspects as price, nutrition label, product appearance, etc. Consumers ranked price first, product appearance and other similar attributes next highest, ingredient list next, and the nutrition label the least. This rank order corresponds almost perfectly with actual usage of these components on Table 4. In general, price and product appearance, etc. (i.e., the "Other" category) were used the most, the ingredient list next, and the nutrition label the least. Unfortunately, consumers were not asked on the initial survey to rate their attention to open date information. However, in pure descriptive terms, consumers' self-reports of food label use and their actual use does correspond.

The use of some components (especially the ingredient list) of the food label appear to be mediated by the education and concern level of the respondents. Education might have influenced whether or not consumers understand what the various ingredients are. However, when the consumers were not concerned with the food products, education had little effect. The consumer's concern could stem from several areas, such as the carcinogenic properties of various ingredients. These consumers might have had health problems or lived with those that do. As a results, they attend to the ingredients.

Why education and concern did not have an effect on nutrition label use and open date use is difficult to explain. Part of the reason is no doubt the fact that these label components do not appear on most products; consequently, in this study only 23 (30%) people used nutrition information and 27 (35%) open dates. And most of these people only used it once during the shopping trip. It is quite possible that further analysis will clarify the determinants of usage of these label components.

A rather simple answer is suggested for the nonsignificant results obtained when price was the dependent measure. Consumers, no matter their education, ethnic background, etc. are always seeking a good price.

In summary, consumers do appear to actually use the information on the food label they report to use. However, how consumers use various components of the overall label is mediated by certain individual differences. There is substantial evidence provided here, but future research needs to be conducted to uncover the nature of these differences and their effects on information use.

REFERENCES

Baumgardner, M. H., Heimbach, J. T., and Stokes, R. C. (1980) FDA 1978 Consumer Food Labeling Survey: Demographic Analysis of Survey Responses. Department of Health and Human Services, Food and Drug Administration, Bureau of Foods, Division of Consumer Studies: Washington, D.C.

Cozby, P.C. (1977), Methods in Behavioral Research. Mayfield Publishing Co.: Calif., 1977.

Heimbach, J. T. (1979), Preliminary Report on the FDA 1978 Survey of Consumer Food Label Awareness, Understanding, and Usage. Department of Health and Human Services, Food and Drug Administration, Bureau of Foods, Division of Consumer Studies: Washington, D.C.

Hackleman, E. C., and Heimbach, J. T. (1980), Food Freshness and Open Dating. Department of Health and Human Services, Food and Drug Administration, Bureau of Foods, Division of Consumer Studies: Washington, D.C.

Hackleman, E. C., and Vandenberg, R. J. (1980), The Relative Value of Different Components of Food Label Information. Department of Health and Human Services, Food and Drug Administration, Bureau of Foods, Division of Consumer Studies: Washington, D.C.

Heimbach, J. T., and Stokes, R. C. (1979). FDA 1978 Consumer Food Labeling Survey. Department of Health and Human Services, Food and Drug Administration, Bureau of Foods, Division of Consumer Studies: Washington, D.C.

----------------------------------------