Consumer Information Processing of Otc Drug Label Information

Joyce L. Grahn, University of Minnesota
ABSTRACT - Government requirements for nonprescription or over-the-counter (OTC) drug labels concentrate on accuracy and appropriateness of information but ignore individual differences in information processing. This paper presents an experimental approach to subjective and objective effects of alternative communication formats on information processing of OTC drug-label information.
[ to cite ]:
Joyce L. Grahn (1980) ,"Consumer Information Processing of Otc Drug Label Information", in NA - Advances in Consumer Research Volume 07, eds. Jerry C. Olson, Ann Abor, MI : Association for Consumer Research, Pages: 633-638.

Advances in Consumer Research Volume 7, 1980     Pages 633-638


Joyce L. Grahn, University of Minnesota


Government requirements for nonprescription or over-the-counter (OTC) drug labels concentrate on accuracy and appropriateness of information but ignore individual differences in information processing. This paper presents an experimental approach to subjective and objective effects of alternative communication formats on information processing of OTC drug-label information.


The issue of information processing in the health care industry is currently both interesting from a scholarly viewpoint and important for public policy development. Dwyer (1978) evaluated cognitive and behavioral impacts for prescription drug therapy. The findings indicated that type of patient package insert had no measurable effects on patient information processing of drug information.

From a public policy viewpoint, information processing of nonprescription drug label information affects large numbers of individuals. It is estimated that within any two day period, over one-third of the population uses an OTC product (Rabin and Bush, 1975; Helm, 1977). Consequences of inappropriate OTC drug use include deprived therapeutic benefits, accidental overdose effects, adverse side effects, and economic waste. Physical risk from misuse can be high. Adverse reactions and accidental overdose are responsible for drug induced illnesses which can require hospitalization and in some cases result in death (Silverman and Lee, 1974; Caranasos et al., 1974; Buchanan et al., 1976; Pediatrics, 1976).

The urgency of the information processing issue for OTC drugs is exacerbated by Food and Drug Administration panels convened to review OTC drugs for safety, efficacy, and appropriate labeling (Moxley et al., 1973). At some time in the future, kinds and amounts of label information will be specified for all OTC drug product categories (FDA Consumer, 1974) and in specific instances the government will dictate label text (Federal Regulations, 1977). Included in the review is the mandate to prescribe labeling that can be understood by persons of "low comprehension" (Federal Regulations, 1977).

Yet low comprehension for the general population has not been defined. Thirty-three percent of the adult population has not graduated from high school (U.S. Bureau of the Census, 1977). Approximately twelve million people fourteen years of age or older cannot read as well as fine average fourth grader (American Education, 1974). However, no attempt has been made to measure either the general population's or these special group's ability to understand OTC drug label information.


The 1978 Engel, Blackwell, and Kollat Model (Engel et al., 1978) lends itself to the investigation and analysis of consumer information processing. Interrelationships between components of the information processing stages and product brand evaluation facilitate experimentation on information processing in the context of a purchase decision. Key components for the experiment in this study are attention, comprehension, and beliefs and attitudes; therefore past research related to these components and relevant for OTC drug label design is examined.

OTC labels often do not attract attention and readership (Knapp, 1974; Blanken, 1976). However, the use of color and graphics (nonverbal signs and symbols) have been shown to improve attention. Salcedo et al.'s study (1971) regarding pesticide labels determined that red is the attention color and that the use of the red skull and crossbones accompanied by "POISON-DANGER" in red has solid theoretical and empirical support. Blanken (1976) found that subjects preferred a readership stimulation symbol (a stop sign) to the same information in narrative form.

Pyrczak (1973) and Palumbo (1974) measured comprehension of OTC drug label information using vocabulary tests. In both studies adult subjects consisted of college students, although Pyrczak also included high school students. Pyrczak and Roth (1976) evaluated readability levels of specific caution and warnings statements on familiar OTC drug labels and gave examples on how to reduce readability levels of specific statements by using the Dale-Chall formula (1948a; 1948b).

Three studies measuring health-related beliefs and attitudes are of interest. Wallston et al.'s (1976) Health Locus of Control Scale (HLC) measures health locus of control beliefs for prediction of health behaviors. Their area-specific HLC measure surpassed a general internal-external measure in explaining behavior. Knapp (1974) reported subjects rated prescription drugs safer than nonprescription drugs and safety ratings for both kinds of drugs decreased from 1966 to 1972. Other findings from this study indicate a minority of consumers lack consensus on drug side effects occurrences.

Findings from a 1969 Food and Drug Administration study (National Analysts, Inc., 1972) indicate that adults are rarely concerned about their health. Most perceive their health to be about the same or better than others their own age. Studies in 1973 (Health United States 1975, 1975), and 1975 (Health United States 1976-1977, 1977) reveal that eighty-seven percent of the population regard their health as excellent or good.

Jacoby et al. (1974a; 1974b) included both objective and subjective measures of information processing in their study of information overload. They concluded (Jacoby et al., 1974a) that subjects feel more satisfied, more certain, less confused, and desire less additional information as the total-amount-of information they have increases even though they make poorer purchase decisions.

No published research investigated the effects on understanding OTC drug label information by simultaneously manipulating readability levels, color, and graphics; nor do any studies investigate information processing across OTC drug product categories. Mediating variables of relevant attitudes and beliefs were ignored in previous OTC label research.


This study investigated alternatives to the present method of information presentation on OTC drug labels in an attempt to facilitate consumer information processing.

Treatment factors for the study were OTC drug product category and level of information presentation. Time and expense associated with label development and sample size limited the experiment to four OTC drug product categories that provided an assortment of indications in the adult general population relieved by using OTC drugs in varying forms. Product categories and their attributes, designated factor A, appear in Table 1.



Subjects received two levels of information presentation --high and low. Low level presentation for each of the four OTC drug categories represented typical OTC drug label information presentation. Each low level label was converted to high level by printing directions and warnings in red to heighten awareness; by simplifying (reducing) readability levels of the directions, warnings, and indications to increase comprehension; and by imbedding a graphic for redundancy in the directions and the warnings sections of the label to heighten awareness and increase comprehension. Information presentation levels, designated factor B, appear in Table 2.



Readability levels on specific label sections were reduced using the Dale-Chall formula (1948a; 1948b). Although grade level would vary as a function of the label, the mean readability level for high level information presentation where readability levels had been reduced was fifth through sixth grade and the mean readability level for low level information presentation was eleventh through twelfth grade.

This study investigated two aspects of comprehension. First, there was concern with subject understanding the messages on the OTC drug label to buy (and eventually use) OTC drugs properly. Understanding was measured objectively by a multiple choice test. Ten items covered each of the seven label areas. Item development presented some unique problems. Because labels were presented in two levels, different comprehension questions were developed to match label information presentation.

Reliability coefficients for the multiple choice questionnaires calculated by Kuder-Richardson Formula 20 (Nunnally, 1967) resulted in a median coefficient of .75. All coefficients were above .55; therefore, the comprehension test does not lack internal consistency; the instrument is reliable.

The second area of interest was the subject's evaluation of the information processing experience. Test items represented were modifications of questions designed by Jacoby et al. (1974b) and evaluated satisfaction and certainty with understanding drug label information and confusion while reading drug labels. The item measuring satisfaction was a seven step scale with anchors from "very dissatisfied" to "very satisfied" in response to the question, "How satisfied are you that you understood the information on the labels?" Certainty was measured by the question, "How certain are you that you understood the information on the labels?" Responses were measured on a seven point scale with anchors from "very uncertain" to "very certain." The final subjective measure evaluated confusion while processing information by the question, "How confused did you feel while reading the labels?" Alternatives on the seven point scale ranged from "completely confused" to "not confused at all."

Because no readability formula can accommodate a specialized vocabulary (Manzo, 1970); it was hypothesized that results of an OTC drug terms test could explain variance in the experimental outcome measures. A matching test of 20 items was developed by randomly selecting words from a list of 50 OTC drug label words that college students indicated were mildly difficult to extremely difficult to understand. While it is possible that word difficulty for college students and for the adult general population may be different, there is no reason to believe that words college students found difficult to understand would be easily understood by general population adults. A matching format was chosen because pretesting indicated that subjects needed variation in instrument structure to reduce potential fatigue and boredom problems. Pretesting revealed that error rates on the test identified words subjects could not define.

A 4 X 2 factorial experiment tested treatment effects of OTC drug product category and information presentation level on each of the dependent variables of satisfaction, certainty, confusion, and comprehension. Each subject participated in all four experiments for a single OTC drug product category.

This study tested the following hypotheses:

1.  Level of information presentation and OTC drug pro- duct category have no effect on subject satisfaction with understanding drug label information.

2.  Level of information presentation and OTC drug pro- duct category have no effect on subject certainty with understanding OTC drug label information.

3.  Level of information presentation and OTC drug pro- duct category have no effect on subject confusion while reading drug labels.

4.  Level of information presentation and OTC drug product category have no effect on subject comprehension of drug label information.

5.  Correlation coefficients are all zero for the following relationships:

a. Understanding OTC drug terms with:

1) Satisfaction

2) Certainty

3) Confusion

4) Comprehension

b. Health Locus of Control with:

1) Satisfaction

2) Certainty

3) Confusion

4) Comprehension

5) Concern for health

6) Understanding OTC drug terms

c. OTC drug use with:

1) Concern for health

2) State of health

6. Beliefs and attitudes regarding the following:

a. Prescription and OTC drugs are equally safe. Correlation coefficients are all zero for the following relationships: OTC drug safety with:

1) Satisfaction

2) Certainty

3) Confusion

4) Comprehension

b. Prescription and OTC drugs are equally effective.

c. Drugs in pill form and drugs in liquid form are equally effective.

d. Prescription drugs and OTC drugs have side effects that are equally severe.

e. Prescription drugs and OTC drugs have accidental overdose effects that are equally dangerous.

Labels were developed for two fictitious products for each of the four OTC drug product categories and for two spray paint to disguise study focus. Subjects were exposed to the same information presentation levels on all labels. That is, each subject received spray paint and OTC drug labels at either the high level or the low legal.

Subjects were 200 adults residing in a metropolitan area from March 15 to May 15, 1978. A quota sample was drawn with control characteristics of age and education similar to the national population. Incentives of $2.00 per completed interview were given.

Data collection occurred at local gathering places in five stages. Each subject was exposed to a screening questionnaire which measured purchase and use of OTC drugs and other consumer products to separate naive and experienced subjects. Only naive subjects who lacked purchase and use experience for the OTC product category under investigation became part of the final sample to minimize the impact of previous acquisition and inventory lng of product-related information. Each subject was then given a large envelope containing three smaller envelopes numbered 1, 2, and 3. Envelopes were opened one at a time beginning with envelope 1. Subjects were first exposed to spray paint labels. They were advised they had a need for spray paint and were asked to read the two labels and make a brand choice decision, place an X on the label they chose, and return the labels to the envelope. Without referring to the labels, subjects answered a set of subjective questions measuring satisfaction, certainty, and confusion. Subjective questions were answered and returned to envelope i before comprehension questions were begun. Completed comprehension questions were then returned to the envelope. The process was repeated with envelope 2 for an OTC drug product category. Each subject was given the set that he or she was experiencing the specific minor health problem contained in their respective stimulus materials. Instructions implied the products would provide relief and subjects were asked to chose one of the alternative products from the label information. (Debriefing indicated that subjects were actively involved in this role play.) Thus, subjects made two brand choice decisions--one for spray paint and one for an OTC drug product. Finally subjects were asked to complete a general questionnaire in envelope 3 measuring beliefs and attitudes regarding drugs; overall personal health; health locus of control; understanding OTC drug terms; and demographic characteristics. Appropriate questions relating to spray paint and oven cleaners were dispersed throughout the questionnaire to continue to disguise research focus. Because of study objectives and experiment length and complexity, subjects were given group instructions. Subjects were allowed as much time as they needed to complete the experiment and were then debriefed. Debriefing averaged fifteen minutes per subject primarily because nearly all subjects had not guessed the focus of the study.


A two-way analysis of variance with 25 observations per cell was used to examine the relationships of information presentation level (hereafter referred to as "level") and OTC drug product category (hereafter referred to as "category") to the four experimental outcome measures. A .05 level of significance was required to reject the null hypothesis. The analysis of variance between each of the subjective score means appears in Tables 3, 4, and 5.







Data supported null hypotheses that manipulations in level and category do not make a significant impact on the subjective experimental outcome measures.

Responses on the subjective outcome measures tended to cluster at the upper levels indicating that subjects were satisfied, certain, and not confused as they processed OTC drug label information. Table 6 provides a comparison of responses in the two most positive alternatives for each of the measures.

Scores for comprehension of drug label information had a potential range of from 0 for no correct answers to 10 for 10 correct answers. Table 7 summarizes results of the analysis of variance by level and category.





Neither level nor category impacted significantly on subject's comprehension scores. The mean score correct for both levels was sixty percent with standard deviations of 2.4 (high level) and 2.6 (low level).

Although this experiment was set up to measure overall comprehension of OTC drug label information rather than separate out the effects of readability levels, color, and graphics; there was some indication that graphics affected comprehension. Distribution of error rates by level for the multiple choice questions is presented in Table 8.



Questions 2 and 3 involved the use of graphics at high level. Error rates were lower for these questions when information was presented at the high level. There was an indication that graphics may have affected information processing on other areas of the label, also. Questions 6, 9, and 10 had error rates of 32%, 64%, and 53% at high level compared to error rates of 23%, 54% and 42% at low level. Thus, findings suggest that graphics improve subject understanding of a particular direction or warning at the expense of subject understanding of brand name, active ingredients, and manufacturer. Question 1, however, had a slightly lower error rate at low level, despite the use of color and simplified readability levels at the high level. In every case, question 1 involved numbers related to dosage directions. It may be that subjects at high levels were distracted by the use of graphics on other directions and warnings or needed symbols to heighten attention and improve comprehension on the non-graphic direction.

Because wording of the comprehension items matched information presentation levels on the labels, there was a potential internal validity problem where items measuring similar information were worded slightly different. Potential internal validity problems would be identified error rates on questions worded exactly alike for all respondents were similar. Questions 6 and 9 were worded the same for all levels and had identical responses within each product category, yet error rates of 32% and 64% occurred with high levels and 23% and 54% with low levels. There is no reason to believe readability level of the questions differentially affected comprehension.

Although subjective measures covaried, the low correlation between subjective measures and the objective measure of comprehension was unexpected. See Table 9.





Simplified readability levels were involved in questions 1, 2, 3, 4, 5, and 8 and error rates were lower at high level for all these questions except 1. Question 1 and questions 6, 7, 9, and 10 had higher error rates at the high level. Age covaried negatively and education and income covaried positively with comprehension. T-tests failed to provide evidence of mean experimental outcome scores by sex of respondent.

Because of the high r2 for comprehension and education, a two-way analysis of variance was used to examine the relationship between label type (level-2 by category-4) and education (collapsed to no college vs. college) to comprehension. Results are presented in Table 11.



Findings indicated that college experience effects comprehension of OTC drug label information. The mean comprehension score for no college was 5.58 correct (out of 10) and for college was 7.47. Fifty-eight students (twenty-nine percent of the sample) had college experience.

The criteria for rejection of the null hypothesis when correlation procedures were used to analyze relationships was .05 level of significance and r2 $.10.

Subjects were able to match 55% of the OTC drug terms with their appropriate meanings. Comprehension of drug label information scores covaried with OTC drug vocabulary scores and explained 31% of the variance in the vocabulary scores.

Health locus of control scores were similar for both information levels. The scores did not covary with the experimental outcome measures, understanding OTC drug term scores, or concern for health.

A matched pairs t-test was used to examine the attitudes regarding prescription and OTC drugs. Data appears in Table 12.



Summarizing the results of significant findings: (1) OTC drugs had mean ratings of somewhat safe, somewhat effective, with somewhat dangerous to dangerous overdose effects and (2) prescription drugs were perceived to be safer, more effective, and have more dangerous overdose effects than OTC drugs.

Data analysis failed to support a relationship between OTC drug safety and the experimental outcome measures.

Associations between personal health and OTC drug use did not meet the required .05 level of significance. Therefore, there was no reason to believe that personal health measures covaried with OTC drug use.

Data analysis of attitude and belief measures indicated there was no evidence of a differential reaction effect from information presentation levels of the experiment.


Findings from this research indicate that labels do not make a difference in information processing--people do. Additional research is needed to explain experimental outcomes. This generalization seems warranted even given the traditional limitations of laboratory research. Subject evaluation of information processing experience may be explained in several ways. Subjects may have made brand choices on very few label items. Thus, subjects may have believed they understood sufficiently to make a brand choice. Subjects were asked to read the labels but maybe they read only relevant areas. Results of group depth interviews indicated subjects read the entire label however. An additional possible explanation for the similar experimental outcome means for both information presentation levels is that the number of items of information per label exceeded the maximum number subjects could process. Findings suggest that information presentation levels influenced which label areas subjects comprehended At low levels, simple items such as brand name, active ingredients, and manufacturer were remembered. At high levels, graphics and color highlighted certain areas and apparently subjects concentrated on and remembered those areas. Thus, a logical explanation is that reduced readability levels, color, and graphics impacted comprehension.

The low correlations between the subjective and objective measures in understanding OTC drug label information were surprising. It was posited that measures to improve comprehension would be sufficiently effective to impact both types of comprehension.

Findings from this study have public policy implications. Consumers do seem to be unaware that they are making decisions based on inadequate or inaccurate information. The problem is especially acute for older people and people with low incomes and limited educations.

Further, if information overload is an issue, then type of information acquisition presents an interesting trade-off--brand and manufacturer vs. directions and warnings for example. Implications are enormous.

Findings of this study, which are not dissimilar to the findings in the Dwyer study (1978), could also be interpreted that because labels do not make a difference in information processing, the Food and Drug Administration should abandon interest in labeling. Resources expended on OTC drug label format and content may be wasted. Future research on other formats could shed light on the problem of optimal label design.

Results of this study raise a number of issues. One relates to graphics' influence on information processing. Another relates to their consumer acceptance--a recently completed study by the Food Marketing Institute (Hammonds, 1978) suggests that specified kinds of graphics on food labels are not well received. Future studies could examine consumer acceptance of types and numbers of graphics per labeling product category.

Another issue for examination is the lack of relationship between subjective and objective measures. Future research could address whether subjective measures are a function of personality of the respondent and comprehension scores a function of ability and experience. Future research could also identify the evaluative criteria consumers use when buying (and when using) OTC drugs. Knowledge of consumer ratings of items of information would aid in understanding information processing.


Blanken, Gary E. (1976), Consumer Attitudes Toward Over-the-Counter Drug Labels, An OPE Study, Food and Drug Administration: n. p.

Buchanan, N.; Cane, R. D.; Glantz, R.; Hunt, J. A. (1976), "Phenolphthalein Poisoning," SA Medical Journal, 50, 1060-61.

"Camphor Poisoning: Over-the-Counter Dangers," Pediatrics (1976), 57, 428-31.

Caranasos, George J.; Stewart, Ronald B.; and Cluff, Leighton E. (1974), "Drug Induced Illness Leading to Hospitalization," JAMA, 228, 713-17.

Dale, Edgar, and Chall, Jeanne S. (1948a), "A Formula for Predicting Readability," Educational Research Bulletin, 27, 11-20.

Dale, Edgar, and Chall, Jeanne S. (1948b), "A Formula for Predicting Readability: Instructions," Educational Research Bulletin, 27, 37-54.

Dwyer, F. Robert (1978), "Consumer Processing and Use of Supplementary Drug Label Information," in H. Keith Hunt, ed., Advances in Consumer Research, Vol. V, Ann Arbor: Association for Consumer Research, 220-26.

Engel, James F.; Blackwell, Roger D.; and Kollat, David T. (1978), Consumer Behavior, 3rd ed., Hinsdale, Illinois: The Dryden Press.

(1977), "Food and Drug Administration," Federal Regulations, 21 (FR 300.1, 300.5), April, 135-138.

Hammonds, Tim (1978), "Nutrition and the American Food System: Part II," a paper presented at the Food Marketing Institute - Family Circle - CNI National Nutrition Conference in Washington, D.C.

(1975), Health United States 1975, DHEW Publication No. (HRA) 76 1232, HE 20 6018:975.

(1977), Health United States 1976-1977, DHEW Publication No. (HRA) 77-1232, HE 20 6018:976-77.

Helm, DeWitt F., Jr. (1977), "Self-Medication: Its Present and Future as a Vital Part of Total Health Care," American Druggist, 175, 38-42, 62.

Jacoby, Jacob; Speller, Donald E.; and Berning, Carol Kohn (1974a), "Brand Choice Behavior as a Function of Information Load: Replication and Extension," Journal of Consumer Research, 1, 33-42.

Jacoby, Jacob; Speller, Donald E.; and Kohn, Carol A. (1974b), "Brand Choice Behavior as a Function of Information Load," Journal of Marketing Research, 11, 63-9.

Knapp, Deanne E., ed. (1974), Consumers and Medication, PB-232 172, Springfield, Virginia: National Technical Information Service, U.S. Department of Commerce.

Manzo, Anthony V. (1970), "Readability: A Postscript," Elementary English, 47, 962-65.

Moxley III, John H.; Yingling, Gary L.; and Edwards, Charles C. (1973), "The Food and Drug Administration's Over-the-Counter Drug Review: Why Review OTC Drugs?" Federation Proceedings, 32, 1435-37.

National Analysts, Incorporated (1972), A Study of Health Practices and Opinions, PB-210 978, Springfield, Virginia: National Technical Information Service, U.S. Department of Commerce.

(1974), "NIE Attacks the Reading and Language Skills Problem," American Education, May, 35-36.

Nunnally, Jum C. (1967), Psychometric Theory, New York: McGraw-Hill Book Company.

(1974), "OTC Drug Review: An Update," FDA Consumer, May 12-18.

Palumbo, Francis Bernard (1974), An Experimental Study of Consumer Expectations and Knowledge of a Hypothetical Over-the-Counter Cold Preparation, Ph.D. Thesis, University of Mississippi.

Pyrczak, Fred (1973), "Reading Comprehension of Directions on Medication," Reading Improvement, 10, 32-33.

Pyrczak, Fred and Roth, Dorothea H. (1976), "The Readability of Directions on Non-Prescription Drugs," Journal of American Pharmaceutical Association, 16, 242-43, 267.

Rabin, David L. and Bush, Patricia J. (1975), "Who's Using Medicines?" Journal of Community Health, 1, 106-17.

Salcedo, Rodolfo N.; Read, Hadley; Evans, James F.; and Kong, Ana C. (1971), Improving the Communication Adequacy of Pesticide Labels: Phase 1 Summary Report, Agricultural Communications, College of Agriculture, University of Illinois at Urbana-Champaign.

Silverman, Milton and Lee, Philip R. (1974), Pills, Profits, and Politics, Berkeley, California: University of California Press.

U.S. Bureau of the Census (1977), Current Population Reports, Series, P-20, No. 314, "Educational Attainment in the United States: March 1977 and 1976," U.S. Government Printing Office, Washington, D.C.

Wallston, Barbara S.; Wallston, Kenneth A.; Kaplan, Gordon D.; Maides, Shirley A. (1976), "Development and Validation of the Health Locus of Control (HLC) Scale," Journal of Consulting and Clinical Psychology, 44, 580-85.