Nutrition Information: a Look At Some Processing and Decision Making Difficulties

ABSTRACT - The study, using simulation techniques, explored the information manipulation complexities and results associated with selecting foods for a day's menu employing (1) an optimizing choice rule using nutrition labeling information, and (2) a food group rule ignoring label data. The findings suggested consumers will experience substantial difficulty in attempting to use current labeling information to buy adequate nutrition at minimum cost.


Zarrel V. Lambert (1977) ,"Nutrition Information: a Look At Some Processing and Decision Making Difficulties", in NA - Advances in Consumer Research Volume 04, eds. William D. Perreault, Jr., Atlanta, GA : Association for Consumer Research, Pages: 126-132.

Advances in Consumer Research Volume 4, 1977   Pages 126-132


Zarrel V. Lambert, University of Florida

[The author wishes to gratefully acknowledge the significant work of Mr. Paul W. Miniard in developing the set of food alternatives with attendant nutrition information and the invaluable assistance of Dr. Gary J. Koehler, Mr. J. Evan Eldridge, Mr. Travis A. Herr, and Mr. Daniel W. Wilmoth in conducting the computer analysis.]


The study, using simulation techniques, explored the information manipulation complexities and results associated with selecting foods for a day's menu employing (1) an optimizing choice rule using nutrition labeling information, and (2) a food group rule ignoring label data. The findings suggested consumers will experience substantial difficulty in attempting to use current labeling information to buy adequate nutrition at minimum cost.


The analysis reported in this paper explored some of the information manipulation and choice difficulties inherent in using nutrition labeling information to minimize food costs and to obtain sufficient amounts of nutrients. Menu decisions were simulated on a computer employing (1) an optimizing choice rule using nutrition labeling information and (2) a much simpler, more traditional rule based on food groups. The quality of the decisions resulting from the two rules were compared. Findings from the exploratory segment of the study suggested consumers may experience extreme difficulty in utilizing nutrition labeling information, in its present form, to select foods having minimal cost that provide adequate nutrition. If this capability is or becomes a goal of public policy, new label indicators of nutrient [For the sake of brevity, the term nutrients in this paper will refer to protein, vitamins A and C, thiamine, riboflavin, niacin, calcium, iron and phosphorus.] levels may have to be developed.

Nutrition labeling is becoming a reality, although current regulations permit more voluntarism on the part of food producers and marketers than some consumer advocates would like (Federal Register, 1973). As time passes, pressure is likely to build to expand the mandatory labeling provisions and to extend nutrition labeling to cover foods that are currently exempted.

Future expansions of mandatory labeling as well as current requirements entail costs that must be paid by society through higher retail prices, taxes, or both. At least part of the expense of measuring nutrients in foods, for example, will be passed along to consumers unless the unexpected happens with the full cost being taken from profit margins. Any measurement activity undertaken by government must be financed by tax dollars. Any additional costs incurred in disseminating nutrient data through product labels, in-store informational displays, and other means may be borne ultimately by consumers. Tax dollars have to be collected to pay for the administration and enforcement of nutrition labeling regulations. Also it seems likely there will be calls for government sponsored educational programs to teach consumers how to use nutrition information listed on product labels.

This brief sketch of some of the costs is not intended to argue against nutrition labeling. It does, however, raise questions about certain potential benefits and whether nutrition labeling can and should be improved so it might be much more beneficial to certain consumer groups. Whether there is a need to make nutrition labeling more usable and thus beneficial, particularly for low income consumers, both the young and the elderly, was one of the issues addressed in part by this exploratory study. Wilkie (1975) has pointed out that societal results from information disclosures are dependent upon consumers' capacities to process the information, and that if information programs are to be successful they must take these and related factors into account. This study was undertaken in the spirit of exploring the capacities required to deal with nutrition labeling information, given one specific purchasing objective.

Several consumer benefit related arguments have been made for mandatory nutrition labeling. One is simply that consumers have a "basic" right to know the nutrient levels of various foods and brands being marketed. Another is that some people with certain dietary and medical problems need to know what is in various foods so they can get the particular nutrients needed and/or avoid ingredients that are detrimental to their well-being. A third reason is that people are often unfamiliar with the nutritional make-up or processed foods. Some of the newer processed foods are at least partially formulated from ingredients that are far removed in the consumer's mind from what the final product resembles.

Another consumer benefit said to result from nutrition labeling is that disclosure of this information will bring competitive pressure to bear on food producers and processors to improve the nutrient levels of their products. Hence, more nutritious foods will become available to consumers, presumably resulting in more nutritional value per food dollar. This rationale assumes consumers will be influenced by the label information to shift their choices toward the more nutritious brands and options.

Still another potential benefit is that nutrition labeling will aid consumers in obtaining the U. S. Recommended Daily Allowance of important nutrients while lowering their food costs if that is their desire. This benefit is a very important one if it can be realized. It can be a major step toward correcting malnutrition among low income families, the elderly with limited resources, and others who may be undernourished partially because of economic constraints. In order to realize this particular benefit, nutritional data on labels must be of a form and quantity that consumers can efficiently manipulate and effectively utilize in decision making. Whether current nutrition labeling is of this nature is one of the questions explored in this study.


What is the complexity and magnitude of the tasks involved in manipulating nutritional labeling information so as to choose food alternatives that provide recommended daily amounts of nutrients at near minimum cost? Is it likely that consumers, particularly the economically disadvantaged and poorly educated, can utilize forthcoming label information to purchase nutritious menus for close to the lowest possible expenditure? In other words, how much difficulty is involved in making optimal choices based on price and nutritional data?

One can take the position that policymakers will never envision the prospect of consumers using nutrition labeling and price information to select nutritious diets at minimal costs. Such a position assumes that extensions to nutrition labeling and tax supported consumer education programs will never be proposed with this goal in mind. If one is comfortable with this position, the first research question may seem highly artificial and academic. In one sense, the first research question is of a "what if" nature. If this was envisioned as a goal, how complex and difficult would the information manipulation tasks be?

A second research question focused on exploring the amount of suboptimality that would result from utilizing a relatively simple, traditional choice rule based on food group classifications rather than directly on nutritional data. The food group approach to menu planning is easy and simple from information manipulation and decision making standpoints, and it has been widely promulgated in schools, cookbooks, and other nutritional guides. The suspicion that it will be extremely arduous to effectively utilize nutrition labeling information in making optimal choices, in terms of sufficient nutrition at close to the lowest possible cost, was one reason for evaluating decisions based on a simpler choice rule.

It might be noted parenthetically that the area of nutrition labeling offers some unique advantages in studying information processing and consumer decision making. The quantifiable nature of nutritional measures and price coupled with the objective standards reflected in recommended daily allowances of nutrients makes it possible to evaluate the quality of outcomes from various information processing and decision making models. Of course, taste preferences have to be dealt with and some other simplifying assumptions have to be made. It is possible, for example, to employ optimizing procedures to select from among alternatives a daily menu that both minimizes total food costs and contains the recommended allowances of all nutrients. Such optimal solutions, then, can serve as a standard for evaluating the quality of choices resulting from the use of various other models.

It is much more difficult in most other consumer choice situations to identify the "best" decision because quantitative measures of product attributes and objective standards are lacking. Consequently, it is hard to assess the quality of consumer choices or the effect of various rules on decision quality.


In exploring the research questions, the information manipulation and choice tasks were narrowed to the specific job of selecting a menu for a single day independently of other days. One day constituted a convenient and appropriate unit of time because nutrition labeling information is stated in terms of U. S. Recommended Daily Allowances. [Hereafter U. S. Recommended Daily Allowances will be called U. S. RDA's.]

A set of alternatives totaling 57 foods and brands, 'shown in Table 1, was devised for possible inclusion in a day's menu. Alternatives were limited to this number for two reasons. A future phase of the study will involve having consumers choose from among the same set of alternatives. Thus, the total had to be limited to a manageable number for field research.



The capacity of available computerized optimizing procedures was the second reason for limiting the number of alternatives. Current regulations stipulate that, given certain circumstances, the label must list U. S. RDA percentages for seven nutrients in addition to protein. For purposes of the current study, the U. S. RDA percentage was included for one additional nutrient which is among the twelve that may be listed on the label at the labeler's option. When price was included, this brought to 10 the number of pieces of information per alternative to be considered in choosing an optimal menu. [This total does not include other nutrition labeling information that states the calorie, carbohydrate, and fat content in absolute quantities instead of percentages.]

The resulting 570 pieces of information (10 times 57 alternatives) is a large number even for a computerized optimizing procedure as subsequent sections will show.

In arriving at an optimal menu, two interdependent problems must be solved: (1) which alternatives to choose; and (2) the quantity that should be included of each chosen alternative. In dealing with the second problem, quantity was measured in terms of servings, the unit used for measuring and stating the U. S. RDA percentages in nutrition labeling. The question of how much of a chosen alternative to include in the menu can be answered in integer or noninteger terms; that is to say, whole or fractional servings.

Integer solutions were sought in this study for several reasons. One is that most consumers probably have been accustomed to think in terms of whole servings. For instance, the food group approach to menu planning is stated in terms of whole servings. Also U. S. RDA percentages in nutrition labeling are for whole servings. Another reason is consumers would probably experience much more cognitive difficulty in attempting to effectively utilize nutritional information in choosing a menu that consisted of fractional rather than whole servings. It would mean consumers would have to recalculate U. S. RDA percentages to correspond to whatever fractional servings, say 1.35 and 1.80, were being considered for the menu. This task appears far more difficult than dividing container size by price which proved so formidable for consumers that calls arose for unit pricing. Since consumers are likely to think in terms of integer or whole servings when making menu choices, the analyses in this study were based on integer solutions. However, in actual practice many consumers may simply decide on how much they wish to eat of a particular food ignoring both the quantity constituting a serving for nutrition labeling purposes and the attendant U. S. RDA information.

Several constraints were imposed on the optimizing and food group choice rules for the purpose of adding realism to the results. Because people typically eat a variety of foods during a day, constraints were placed on the total number of servings that could be chosen of any one alternative food. Otherwise the daily menu resulting from the choice rules might have contained an unrealistically large number of servings of only one or two foods that happened to he high in nutrients and low in cost. Since people usually do not mix brands of the same product, say canned peas, by eating one serving of a certain brand and another serving of some other brand in the same day, multiple servings of a product were constrained to a single brand. These and the other constraints are shown in more detail in Table 1.

Of course, any constraints of this nature contain a degree of arbitrariness since what one envisions as a reasonable or typical menu is a function of personal taste and custom. Consequently, some readers may feel the constraints imposed in the study should be modified in one way or another. However, modified constraints might also be questioned by other individuals, so perhaps unanimity would be unlikely for any set of constraints. Although the particular constraints imposed here could be changed in any number of ways, they are more or less reasonable and thus can be used in generating results of different choice rules for comparison purposes.

The 57 food alternatives were classified into the four basic food groups that people have been advised to eat each day for good nutrition: bread and cereal, fruit and vegetable, milk, and meat (for example, see Schaupp and Schaupp, 1975). Then the food group choice rule was applied by randomly selecting a specified number of servings from each group. Random selection, within the constraints described earlier, was used because there was no convincing rationale for choosing one alternative over another within a group. Some choice rule might be devised which incorporates criteria for selecting alternatives within a group. But such a rule would not be -characteristic of the traditional food group principle to menu planning. Therefore, no attempt was made within the scope of this exploratory study to devise a new rule. The traditional food group principle does not consider directly the nutrient levels of alternatives within a group. Thus, it essentially ignores nutrition labeling information.

A total of 60 daily menus were chosen with the food group choice rule. A sampling of results was obtained to lessen the possibility of getting atypical menus that were either very high or low in nutrition and cost. The sampling of results was used to address the question of how much suboptimality would occur from employing a simple, rather than an optimizing, information manipulation and choice rule.


The findings are presented and discussed in four sections. The first one looks at the complexity of making an optimum choice of alternatives, and the second at some results of an optimum choice rule. The third section presents the results of the food group choice rule and shows comparisons with the optimizing rule. The final section describes the worst choice of alternatives in order to give added perspective for evaluating the quality of decisions resulting from various information manipulation and choice rules.

Task Complexity of Optimum Choice

In spite of a computerized procedure, it proved to be a lengthy and laborious task to analyze price and 9 pieces of nutritional information per alternative so as to select a menu that contained the U. S. RDA's for all 9 nutrients at a minimum total cost. It took approximately 450 seconds of CPU time on an IBM Model 370/165 system to arrive at an optimal solution. In other words, about 7 1/2 minutes of computer time were required when the number of alternatives was limited to 57 which is only a small proportion of the number confronting consumers in the typical supermarket. Probably around 6000 alternative foods and brands are available in today's supermarket. If all had nutrition labeling showing U. S. RDA percentages for protein and the seven vitamins and minerals that currently must be listed, consumers would be faced with 54,000 pieces of information, including price, to consider in making near optimal choices. The complexity and difficulty of the information manipulation and choice tasks that would be involved are nearly unfathomable.

Admittedly the simulated purchasing problem posed in this study was a very rigorous one, and perhaps one which consumers would not attempt to solve. The solution, which was called for, required the 9 pieces of nutrition information per alternative to be considered along with price with the final choice of alternatives providing at least 100% of each U. S. RDA at a minimum total cost for the day's menu. One might argue, based on experience in countries where malnutrition is rampant and programs have been undertaken to change eating habits to include unfamiliar but highly nutritious foods, as well as on other grounds, that few people will override taste preferences and cultural patterns and attempt to substantially lower their food costs by carefully considering nutrition labeling data. The present study looked at the difficulties that would be encountered if consumers did, in fact, attempt or were encouraged to attempt to achieve this objective in their purchasing.

The optimizing procedure utilized in the analysis to identify the "best" alternatives for the day's menu was not the most efficient possible from the standpoint of computer time consumption. A noninteger linear programming model was adapted and employed rather than an efficiently programmed integer model. [A noninteger solution which is the most efficient from a computer programming standpoint required 3.96 seconds of CPU time.] Nevertheless, if 3, 1, or even 1/3 of a minute of high speed computer time is required to make optimal choices from among 57 alternatives, is it plausible to think consumers can deal with nutrition labeling information on a much larger number of alternatives and come close to making optimal choices? To do so they would have to sum across alternative choices the percentage of each U. S. RDA that each alternative contained while making cost trade-offs among nutrients to arrive at the minimum possible total cost for a day's menu which contained 100% of each U. S. RDA. The human mind is a marvelous device for dealing with complexity and a multitude of data, but one must wonder if the mind can deal effectively with a problem of this magnitude.

Consumers who need most to minimize their food costs and at the same time obtain adequate nutrition are probably the least equipped to effectively manipulate nutrition labeling information in its present form. Low income consumers often have poor educations and underdeveloped analytical skills. The elderly living on Social Security or other limited pensions are likely to be ill-prepared to deal with the information in a near optimal manner.

It is not the intent of this paper to argue that nutrition labeling in its present form should be abandoned. Its costs to consumers and society in general may be justified in terms of various other consumer-related benefits (for example, see Deutsch, 1975). The point being expressed is that some additional steps probably have to be taken if label information is to be very effective in assisting the most needy consumers in buying adequate daily nutrition at close to minimum cost.

Innovations like nutrition labeling usually lead to calls for private and government sponsored educational programs to teach consumers how to use the new information. The extreme arduousness of the tasks inherent in an optimal choice procedure, as suggested by this exploratory analysis, raises doubts about the wisdom of educational programs whose goal is to teach consumers how to utilize nutrition labeling to make near optimal menu selections in terms of cost and several nutrients. Before any such educational programs are launched, it should be clearly demonstrated in pre-tests that resulting improvements in consumer decisions justify the costs of the educational endeavor. Perhaps avenues other than education would have significantly better cost-benefit ratios. This possibility certainly should be investigated before substantial funds are committed to educational programs.

One conceivable avenue would be to develop nutritional codes or indices that would collapse the nutrient values of a food into a one or perhaps two dimensional measure. The idea is to have on the label a nutritional measure or indicator which is sufficiently simple from an information processing standpoint that it can be easily and effectively utilized by those consumers who have the greatest need to improve the quality of their food buying decisions.

This idea is expressed with the foreknowledge that many nutritionists will argue it is impossible to combine the values of disparate nutrients into a one or maybe two dimensional indicator without incurring a large loss of nutritional information. Such an argument undoubtedly has considerable validity at the present time. However, two important empirical questions have yet to be answered. One is whether research can lead to the development of a code or index having a smaller loss of information? The second question is whether a simple and more easily used index would be superior, despite its attendant loss of information, to current nutrition labeling in enabling consumers to lower their food costs and at the same time improve their nutritional intake. A simpler code could be listed on the label as an addition to current information and not necessarily as a substitute.

Another course of action would be to distribute to, say, low-income and elderly consumers a sizable number of preplanned daily menus that provide necessary nutrition at near minimum expense. The number of menus might be large enough so several weeks could pass without necessitating the repetition of a menu. Consumers who wished to do so could follow these preplanned menus. One problem, of course, would be in providing consumers with guidelines for choosing the least costly brand of a menu item, but one which contained the necessary amount of nutrients.

Optimum Choices

The optimizing procedure generated two menus that were only two cents apart in total cost. The two menus differed very slightly in calories and nutrients as can be seen in Table 2. The menus were the same foodwise except for two vegetables. The less expensive one contained one serving of sweet potatoes and one of onions. The other menu had two servings of sweet potatoes and no onions. The common contents were eggs, bread, hamburger rolls, macaroni and cheese, carrots, cabbage, and spinach.

The total cost of the chosen items in the least expensive menu was 92.8 cents. The cost ran about .00055 cents per calorie.

Cost in relation to nutritional value was computed by dividing two different U. S. RDA indices into total menu cost. The resulting amount represented a cost per nutritional unit. The principal purpose of these indices as well as cost per calorie figures was to provide a common basis, like unit pricing, for comparing costs since both nutritional levels and total costs varied among menus.

The first U. S. RDA index assumes nutrients in excess of 100 percent of the U. S. RDA's are unneeded and hence should not be included in cost per nutritional level computations. The index is the sum of the U. S. RDA's, expressed as a percentage divided by 100, in the menu with the value of each U. S. RDA limited to a maximum of 100 percent or 1. If, for example, the menu contained 450 percent of the U. S. RDA for vitamin C, only the first 100 percent would be counted in the index. Hence, the total value of the index was limited to 9 in this case since 9 nutrients were considered. Hereafter, this index will be described as bounded.

In contrast to the bounded index, it can be assumed that nutritional levels above 100 percent of the U. S. RDA's are valuable and hence should be included in the computations. Under this assumption, the total index value is the sum of all U. S. RDA percentages in the menu divided~ by 100. If the menu contains 450 percent of the U. S. RDA for vitamin C, for instance, a value of 4.5 is added in the index. Unlike the bounded index, this one does not have an upper limit and can be called unbounded. [If a RDA index proves to be a useful measure of nutritional costs, a more refined index, including both bounded and unbounded components, might be devised in which nutrient levels above the U. S. RDA are included where excess nutrition is valuable and disregarded for those nutrients where it is not.]



The cost per U. S. RDA using the bounded index was 16.4 cents for the least expensive menu. With the unbounded index, the cost fell to 6.8 cents.

By way of comparison, corresponding costs for the more expensive menu were 16.8 and 6.2 cents. Cost per calorie was .00054 cents. The lower unbounded U. S. RDA index and calorie costs were due to higher calorie and nutritional levels of this menu.

It might be noted parenthetically that the number of calories provided by the two menus is approximately that needed on the average by males and females between 23 and 50 years old. The number varies with size of the individual and the amount of physical activity. Pregnant and lactating females typically require more calories and older people less. The amounts of vitamins A and C are within what many accept as safe levels, although there is controversy among experts on what the safe and proper levels are (Deutsch, 1975).

Food Group Choice Rule

How much suboptimality would be incurred from using a simple, widely known choice rule rather than the arduous optimizing one? Stated differently, would the food group rule lead to choices that would be "good enough" since an optimizing choice rule might be nearly impossible for consumers to implement? These were the primary questions addressed by this phase of the analysis.

The two optimal daily menus contained a total of 15 servings; 2 from the meat group, 5 from bread, 2 from milk, and 6 from the fruit and vegetable group. The food group rule was programmed to select these same quantities so that any differences in cost and nutritional levels would be due to the choice rule and not to differences in quantity levels.

The food group choice rule generated menus that contained, when averaged over the 60 days, larger amounts of 6 of the 9 nutrients than the optimal menus. The food group menus were within 6 percentage points of the optimal ones on two of the remaining nutrients. Only in the case of one nutrient, calcium, did the food group menus on the average fall below the U. S. RDA. [This statement is based on the sum of the nutrient levels in the selected menus and the initial starting menu.] This nutritional deficiency was at least partially an artifact of the available alternatives. They were all very low in calcium, ranging from 0 to 10 percent of the U. S. RDA.

Although average nutritional levels were quite adequate, as seen in Table 3, there were large day-to-day variations. The extent of these variations were reflected in the standard deviation, range, minimum and maximum values which are also listed in Table 3.

Another measure of the variation is the number of days when the menu was deficient in one or more nutrients. In contrast to average nutritional levels which satisfied the U. S. RDA's in all but one instance, 15 menus or 25 percent of the total were deficient in at least one nutrient other than calcium. Some menus were deficient by only 2 or 3 percentage points. Others had insufficient levels of two or more nutrients.

The average total cost of the menus selected by the food group rule was $2.52 per day. This figure was 2.72 times higher than the $0.928 cost of the optimal choices. Total cost also varied considerable from day-to-day. The lowest cost was $1.98, which was 2.13 times greater than the optimum. The most expensive menu was $4.42 or 4.76 times the optimum.



Calories and the U. S. RDA indices provide some common points of comparison with the optimal menu since nutritional levels and calories as well as cost were higher on the average for food group menus. Cost per calorie ran .00108 cents which was about 2 times the optimum. The bounded U. S. RDA index cost was 44.7 cents as compared with 16.4 cents for the optimum. Cost per unbounded U. S. RDA was 14.6 cents or 2.15 times higher than the optimal menu.

In summary, the food group choice rule which is simple and easy to use from an information manipulation and choice standpoint may result in adequate nutrition, if substantial day-to-day variations can be tolerated. But the findings of this exploratory study suggest the rule leads to choices that are inefficient from a cost viewpoint. Food expenses under the rule ran 2 to 3 times higher than costs for the optimal choices.

These evaluations measured suboptimality on a unipolar scale, distance from the optimum. However, the quality of decisions under various choice rules can be measured at least conceptually on a bipolar scale; one which reflects the feasible range in decision quality from the best to the worst possible. Results of a particular choice rule, say the food group for example, conceivably could be some distance from the optimum and yet be located relatively far out toward the "best" end of the range in decision quality. In order to assess the food group rule in relative terms and to have a benchmark for evaluations of consumer decisions in future stages of research, the 57 alternatives were analyzed to determine the worst possible choices from the standpoint of cost. [The worst possible from a nutritional standpoint is open-ended and thus subject to innumerable definitions.]

Worst Possible Choices

Two of the constraints imposed on the search for the worst choice should be mentioned. One was the choices had to satisfy the U. S. RDA's for all nine nutrients. The second was a limitation on the overall amount of food to 15 servings, the same as in the optimal menu. The purpose of these constraints was to make the best and worst menus comparable in terms of minimum nutritional levels and total quantity of food. If a constraint was not placed on overall quantity, for instance, much of the difference in cost between the two menus would be solely a function of quantity differences and unrelated to choice rules.

Total cost was $6.46 for the menu consisting of the worst choices from the standpoint of dollar outlay. This was 6.96 times higher than the optimum and 2.56 times greater than the average for the food group choice rule.

Cost per calorie was .00231 cents compared to .00055 for the least costly menu, a difference of 420 percent. The worst menu was 2.13 more costly per calorie than the average for the food group rule.

The bounded and unbounded U. S. RDA index costs were $1.069 and $0.218 respectively. These were 6.25 and 3.21 times greater than those for the optimal menu, and 2.39 and 1.49 times higher than the food group averages.

Viewed in relative terms on a bipolar scale, the food group average menu costs of $2.52 were only 28.78 percent of the distance away from the optimum toward the worst end of the scale. However, the costs under the food group choice rule for the hounded and unbounded U. S. RDA indices appeared much more unfavorable when placed on the bipolar scale. They were 49.39 and 97.33 percent, respectively, of the distance away from the optimal end. One reason for these high relative U. S. RDA costs is that the worst menu had very high levels of most nutrients. This tended to depress the U. S. RDA costs particularly for the unbounded index.

Another measure might involve comparing current food costs for specific groups of consumers with the cost of a so-called optimal menu to ascertain the amount of sub-optimality each consumer group was incurring. In order to do this, the Dumber of alternatives available to the procedure selecting an optimal menu probably would have to be expanded substantially so the universes from which consumer choices and optimal choices were made would be much more comparable. Perhaps acceptable levels of sub-optimality could be defined with food expenses within these ranges being viewed as not being a societal problem provided U. S. RDA levels were being satisfied.


Although this exploratory study has many limitations, the findings reinforced an a priori suspicion which has important societal implications. The complexities and difficulties of the information manipulation and choice tasks necessitated by nutrition labeling would appear to preclude the most needy consumers from using this information to minimize their food costs. If one intent of public policy is, or becomes, to assist low income consumers in buying nutritious menus at minimal cost, then it seems likely that new and additional label indicators of nutrition levels will have to be developed. These indicators should be of such a nature that they can be effectively utilized by the poorly educated and the elderly. Of course, there is the possibility that these consumer groups will not substantially alter their dietary habits regardless of the availability of easy to use nutrition and cost indicators.

Any effort to develop new indicators should draw on knowledge of human information processing capabilities so the resulting indicators could be easily and effectively utilized by those consumers who most need to improve the quality of their food purchasing decisions. Such indicators might be in addition to current nutrition labeling information. Thus, they could serve a complementary role, not necessarily a substitutional one on food labels.

The suggestion of simpler nutritional information indices might be interpreted by some as a leap into the "can too much information be detrimental argument" (Jacoby, 1974; Russo, 1974; Summers, 1974; Wilkie, 1974). This is not the intent of this paper. It is suggested, however, that it maybe too much to expect consumers, particularly the undereducated and elderly, to effectively deal with all the information on nutrition labels in the present form. At the same time, the volume and form of the current information may serve some consumer purposes extremely well.

The particular set of alternatives on which the analyses were based constitute one of the many limitations of this study. The number of alternatives was limited and consisted of only a very small proportion of those available in a typical supermarket. Furthermore, a degree of arbitrariness existed in which alternatives were included and which were not. It easily can be argued that some alternatives should have been omitted and others added. For example, alternatives like fresh fruits and vegetables were included for which nutrition labeling is not currently required.

It can be contended that some of the constraints placed on the choices were inappropriate. Some might argue that multiple selection of alternative brands of the same food should be permitted. Others might assert that too few or too many servings were allowed for certain foods.

One of the most glaring limitations is that the choices were simulated, and data were not collected from consumer subjects.

The imposed task of selecting a minimum cost menu which satisfied nine U. S. RDA specifications was an extremely demanding one, and it may be viewed as a major limitation, particularly if one believes that most consumers can select menus that are within acceptable ranges of cost and nutrition. Consequently, some may contend the optimal task requirement leads to an overly pessimistic picture being portrayed.

One could continue to list limitations, but it is our belief that additional research of a more sophisticated nature will tend to confirm the findings of this exploratory analysis concerning the difficulty some consumers are likely to experience if they attempt to use current nutrition labeling information in purchasing nutritious menus at near minimum cost.


Ronald M. Deutsch, Nutrition Labeling (Bethesda, Md.: National Nutrition Consortium, 1975).

Federal Registrar, 14 (January 19, 1973), 2125-2132.

Jacob Jacoby, Donald E. Speller, and Carol A. Kohn, "Brand Choice Behavior as a Function of Information Load," Journal of Marketing Research, 11 (February, 1974), 63-69.

J. Edward Russo, "More Information Is Better: A Reevaluation of Jacoby, Speller and Kohn," Journal of Consumer Research, 1 (December, 1974), 68-72.

Dietrich Schaupp, and Frederick Schaupp. Consumer Handbook for Senior Citizens (Morgantown, WV: Bureau of Business and Economics, 1975).

John O. Summers, "Less Information Is Better?" Journal of Marketing Research, 11 (November, 1974), 467-468.

William L. Wilkie, "Analysis of Effects of Information Load," Journal of Marketing Research, 11 (November, 1974), 462-466.

William L. Wilkie, How Consumers Use Product Information. A report prepared for National Science Foundation, Research Application Directorate (RANN), 1975.



Zarrel V. Lambert, University of Florida


NA - Advances in Consumer Research Volume 04 | 1977

Share Proceeding

Featured papers

See More


Dimming the Light Offers A Creative Lens: The Impact of Ambient Illuminance on Creativity Assessment

Chen Wang, Drexel University, USA
Ravi Mehta, University of Illinois at Urbana-Champaign, USA
Rui (Juliet) Zhu, Cheung Kong Graduate School of Business, China
Jennifer Argo, University of Alberta, Canada

Read More


The Experiential Advantage: A Meta-Analysis

Evan Weingarten, University of California San Diego, USA
Joseph K Goodman, Ohio State University, USA

Read More


Understanding the Framing of Recommendations

Jia Gai, Erasmus University Rotterdam, The Netherlands
Anne-Kathrin Klesse, Erasmus University Rotterdam, The Netherlands

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.