Printer Friendly

Warranties: continued readability problems after the 1975 Magnuson-Moss Warranty Act.

A warranty is a promise by a manufacturer or a seller to stand behind its products. The purpose of the Magnuson-Moss Warranty Federal Trade Commission Improvement Act (MMWA)--passed by Congress in 1975--is to help consumers before the purchase to comparison shop and get the best warranty and after the sale to require companies to keep their warranty promises. A major thrust of the act was to assist consumers in the purchase decision by making warranties "easy to read and understand. They must be written in ordinary language, not 'legalese'" (Federal Trade Commission not dated, 1).

For sellers of products costing more than $15 that have a written warranty, the MMWA specifies that the warranty must be "full" or "limited" and available for customer review before the sale. A full warranty means that there are no or minimal charges for repair or replacement of the product during the warranty duration. A limited warranty specifically limits seller obligations and places more financial responsibility on the consumer (Wish, Steely, and Tritten 1978). From the manufacturer's perspective, a warranty may serve as a contractual limitation on what the manufacturer is required to do when the product fails within a specified time. For a consumer, a full warranty may require less reading while a limited warranty may require more reading (i.e., consumers must examine the warranty for the manufacturer's restrictions or exclusions for product repair or replacement).

Research has shown that warranties do have a role in consumers' purchase decisions. Wilkes and Wilcox (1981) found that respondents perceived the product with a limited warranty to be less desirable than the product with the full warranty. Other studies found that a product with a superior warranty is associated with greater quality (Olson and Jacoby 1972) and less risk (Bearden and Shimp 1982; Perry and Perry 1976). Wiener (1985) studied appliance and auto warranties and found that consumers perceived warranties as a signal of product reliability.

Prior to the Warranty Act, warranty wording was more appropriate for lawyers than consumers and the "fine print" was deceptive (Wilkes and Wilcox 1976). Early post-act research using readability measures indicated warranties continued to be difficult to comprehend (Shuptrine and Moore 1980) and warranties remained complex from the consumer's point of view (Consumer Reports 1984).

The purpose of this study is to examine the readability and understandability of recent warranties for a sample of durable products. In addition, the mean educational levels required to read and understand full and limited warranties are compared. Have manufacturers and sellers responded to the consumer's desire for warranties written in easy-to-understand terms as MMWA requires?


Two measures of readability were selected to examine warranties. These two measures were the Flesch Count and the Fog Index.(1) The Minnesota Interactive Readability Approximation Program (MNIRAP), a microcomputer program, was used to run the two measures of readability (MNIRAP 1977). MNIRAP can determine readability levels (first grade to professional) by seven different methods.

The Fog Index and the Flesch Count were chosen as the readability measures to examine warranties based on ease of use, reliability, and significant convergent validity results from research (Lehman et al. 1982; Shuptrine and Lichtenstein 1985; Shuptrine and Moore 1980). Spearman rank correlation coefficients using the Fog Index and the Flesch Count measures were significant at the .05 level or less in two studies on warranties and textbook readability (Shuptrine and Lichtenstein 1985; Shuptrine and Moore 1980).

The authors sought to confirm the convergent validity of the two measures found in the Shuptrine and Lichtenstein (1985) study and make approximate comparisons with the Shuptrine and Moore (1980) findings. Also, the Fog and Flesch measures were found to be better measures than the Dale-Chall Method and the Smog Grading Method when evaluating written material that may go beyond the sixteenth grade level (Shuptrine and Lichtenstein 1985).

An example of the calculations used to obtain the educational levels in the two methods is provided in the Appendix. Each measure provides a standardized index value or educational level (number) required for ready comprehension of the written material. The Fog Index is the simpler of the two measures in determining readability (Gunning 1952). The formula is based on average sentence length and difficult words per 100 words (words having three or more syllables). The calculated Fog Index number gives the approximate grade level required for comprehension of the tested material--grade school level through the college graduate level or approximately grade one to beyond grade 16. The grade level calculated for written material is the grade at which people can give nine out of ten correct answers on questions from the tested passages or 90 percent correct comprehension (Gunning 1952).

The Flesch Count formula is a similar test that develops reading scores which depict understanding for the typical or average person at that grade level (Flesch 1951). The procedure uses average sentence length, number of syllables per 100 words, and a correction for personal nouns to determine a readability grade level. The Flesch Count score provides a measure of grade level understanding from grade five to beyond 16.

The Fog Index requires 90 percent comprehension on tested material whereas the Flesch Count represents understanding of the material by the average person (65-70 percent of the population). Therefore, on the same material, the Fog Index grade levels usually average one to two grades higher than the Flesch Count (Shuptrine and Lichtenstein 1985).

The Fog Index and the Flesch Count were used to determine the educational levels required of expected readers of warranties for small and large durable products. National census data were used for comparison with the educational levels estimated using the two tests. In 1987, census data indicated the median number of school years completed for persons 25 years of age or more, whites and nonwhites, is 12.7 years (U.S. Bureau of the Census 1989). The median number of school years completed does not necessarily correlate with people's reading levels. Kozol (1985) reported 50 percent of American adults are not able to read an eighth-grade level book. Consequently, written warranties for products for the general population should approximate the eighth-grade reading level or, at maximum, should not exceed the reading and understandability for 12.7 years of education.


Copies of the product warranties used in the study were obtained in Spring 1988. The ten product categories chosen for warranty study were available at retail stores located in a southeastern metropolitan city with a population of about half a million. Although a random sample process was not used, representativeness was attempted by having investigators visit and collect warranties from a large and diverse group of department stores, discount stores, large and small appliance stores, and bicycle and automobile dealers.

Readability levels were computed for 121 warranties in ten categories of consumer durables: automobiles, bicycles, can openers, coffee makers, food processors, hair dryers, lawnmowers, microwave ovens, radios, and televisions.(2) Product categories were chosen to represent a range of prices and consumer complaints, e.g., automobiles and can openers. Buyers of the products represented by warranties used in this study reflect a diverse and extremely broad, heterogeneous group of consumers. Five of these product categories (automobile, bicycle, can opener, coffee maker, and television warranties) are the same as used by Shuptrine and Moore (1980) and are used to compare data over time. From eight to 15 warranties were analyzed for each product.


Spearman rank correlations were used to indicate if the Fog Index and Flesch Count scores were significantly related to establish a measure of convergent validity of comprehension and readability of the two measures in the ten product categories. All correlations for the observed values are significant at the .01 level indicating the two measures yielded approximately the same results. Level of education needed to comprehend the warranties was consistently less for the Flesch Count than the Fog Index. The Flesch Count for the 121 warranties averaged about two to two and one-half years less in education. These findings are consistent with previous studies and expected. Because the Fog Index requires 90 percent comprehension and the Flesch Count requires 65-70 percent comprehension of the same written material, the expected grade levels for the Fog Index are usually one to two years greater than the Flesch Count. These results, therefore, support the findings from previous research of convergent validity of the two measures and that the two measures are consistent with the degree of comprehension measured with each technique (Shuptrine and Lichtenstein 1985; Shuptrine and Moore 1980) even though the exact same warranties were not examined in previous studies.
Spearman Rank Correlations for Product Warranties Using the Fog
Index and the Flesch Count
Product Category Rs P
Automobile .95 .000
Bicycle .86 .000
Can Opener .64 .007
Coffee Maker .90 .003
Food Processor .99 .000
Hair Dryer .91 .000
Lawnmower .73 .003
Microwave .84 .002
Radio .90 .000
Television .88 .000

Readability of Warranties

Table 2 provides measures of needed educational levels for comprehension of all warranties using Fog Index and Flesch Count measures. Grade level categories, e.g., high school, some college, were determined by the authors for interpretation purposes. The majority of warranties examined required some college education (13-15 years of education) for readability. The Fog Index and the Flesch Count scores, 90.9 percent and 79.4 percent, respectively, indicate that people need 13 or more years of education to read and understand the warranties. Given that the median number of school years completed by persons 25 years or older is 12.7 years, it is evident many people would have difficulty reading and understanding the warranties examined.

Table 3 shows the ranges and means of the two readability measures and the corresponding educational levels needed for each of the ten product categories. Mean scores are used for discussion. For the Fog Index, all ten product categories measured above 15.6 years needed for comprehension with seven of the ten product categories at college graduate level or higher (|is greater than or equal to~ 16.0). The easiest warranties to understand by either method in all product categories (radio) were measured at 13.8 years or greater, i.e., at least some level of college TABULAR DATA OMITTED TABULAR DATA OMITTED study may be required to read and understand all the warranties examined. Can opener, coffee maker, hair dryer, microwave, radio, and television warranties measured from 13.8 to 14.5, some college on the Flesch Count; automobile, bicycle, food processor, and lawnmower warranties measured from 15.2 to 15.4, not quite college graduate level.

Products do not have to be complex or expensive to have more difficult to understand warranties (can opener warranties were more difficult than television and microwave warranties on both measures). The average educational level for all warranties in all product categories was 16.9 years for the Fog Index and 14.5 years for the Flesch Count. Notably, most small and large appliance warranties are not easy to read by most people. All waranties examined in this study clearly would not be easy to read or understand by the 50 percent of the American adult population unable to read at an eighth-grade level. However, as the ranges indicate, there were some product categories where a few warranties (11.6 percent) were written at a high school level or less (Table 2 reports 9.1 percent of warranties using the Fog Index and 20.7 percent by Flesch Count).(3) Perhaps some product manufacturers are trying to differentiate their offerings or increase market share with easier to read warranties.

Full Versus Limited Warranties

Wilkes and Wilcox (1981) suggested that many manufacturers were likely to change from full warranties to limited warranties because of confusion in the provisions, difficulties in interpretations, and the prospect of the cost and burden of the rules and regulations of the MMWA. This change obviously was not an intent or objective of the MMWA. Table 4 indicates in this study that, indeed, more than twice as many limited versus full warranties were provided by manufacturers, 74 and 31, respectively. Warranties for automobiles, lawnmowers, and televisions, the more expensive products, were not full warranties.

Table 5 presents the results comparing the mean educational level needed to read the full and limited warranties used in this study. The results reported in Tables 4 and 5 also are used for two additional comparisons. First, these findings may support the Wilkes and Wilcox's (1981) report of a 1978 survey of manufacturers that they may be switching to limited warranties. In the current study, 2.4 times as many limited as full warranties were offered. Second, the limited warranties required higher educational levels of readability than full warranties: 17.5 versus 16.1 for the Fog Index TABULAR DATA OMITTED TABULAR DATA OMITTED and 15.1 versus 13.7 for the Flesch Count. T-tests were computed for the limited and full warranties using each of the measures. Limited warranties required a significantly higher educational level than full warranties for understanding on both measures--Fog Index t = 2.19 (p |is less than~ .05), Flesch Count t = 2.86 (p |is less than~ .05).

Comparison to Earlier Warranty Research

To provide greater insight, some of these findings were compared to previous research. One-half or five of the product categories (60 warranties) used in this study had been used by Shuptrine and Moore (1980). The average educational levels were compared for warranties from these product categories: automobile, bicycle, can opener, coffee maker, and television. Table 6 indicates warranties measured by the Fog Index for bicycles and can openers increased in reading difficulty over time from some college to college graduate level. Television warranties also increased in reading difficulty level by one full year but remained in the some college category for understanding. Coffee makers decreased in reading difficulty level from the college graduate level to some college. Warranties for automobiles remained in the post college graduate level but decreased by 1.9 years.

Educational levels for warranties examined by the Flesch Count in the 1980 study were reported by educational level category only. The results within a level may be higher or lower; however, all grade levels in the present study are at the some college level. The some college TABULAR DATA OMITTED category represents a decrease for automobile warranties from the college graduate reading level and an increase for can opener warranties from the junior high school level. Overall, warranties in both studies had reading difficulty scores beyond the high school level. While this evidence is not conclusive, the comparisons suggest that warranties continue to require some college or college graduate level for understanding.


The consumer believes, in general, that a warranty should reduce post-purchase risk if the product does not function properly. Yet, the consumer may not have the ability to fully comprehend the actual meaning of the language and "fine print" of a warranty written in legal terminology or "legalese" as termed by the Federal Trade Commission. The results of this study provide evidence that the level of education needed to read and understand most warranties is beyond what one-half of the American population have attained--the 12.7 median years of completed education for adults 25 years or older and much beyond the level of 50 percent of the adult population unable to read at an eighth-grade level. Consumers who do not understand their warranty coverage most likely are less able to comparison shop for the best coverage and obtain all the benefits provided by a warranty.

Given that a large majority of warranties in this study and in the Shuptrine and Moore (1980) study are written at the high school readability level or above (|is greater than or equal to~ 12.0), these findings should prompt the FTC to re-examine the effectiveness of the MMWA or enforce the law as it stands. Since 1975 when Congress passed the MMWA, readability levels of warranties have not significantly changed. Clearly, the MMWA has not accomplished its goal of making warranties easy to read and understand.

In fact, some evidence in this study suggests that the MMWA may have had an opposite effect by increasing the number of limited warranties which tend to have more difficult reading levels. Could it be that more "legalese" is being used because the MMWA has created some confusion for manufacturers as suggested by Wilkes and Wilcox (1981)? Or are manufacturers reducing their risk and providing less warranty coverage? Warranties appear to be written in a language to protect the seller and not the buyer. Consumers prefer a product with a full warranty over a product with a limited warranty (Wilkes and Wilcox 1981), and consumers believe warranties reflect a product's reliability (Wiener 1985). Perhaps companies are missing a strategic marketing opportunity to offer the consumer an easy-to-read, full warranty. Some retailers now provide very simple, no questions asked return policies. Such retail action, however, may shield manufacturers and their warranties from claims and not provide any real incentives for companies to use easy-to-read warranties. Wilkes and Wilcox (1981) showed that a passive presentation of warranty coverage can seriously impair brand evaluation by customers. Because industry leaders often set policies and procedures which other firms follow, an opportunity may exist to gain market share advantages by being the first to offer consumers the best, easy-to-understand warranty.

From a public policy viewpoint, the MMWA apparently is not performing as intended based on results of this study and previous research. The FTC should recognize this problem and implement a division to work with companies to make warranties easier to read and understand by the average citizen. The findings of this study and previously cited studies should be generalizable to most warranties. The implication is that at a minimum, the FTC should require about a twelfth-grade readability level and consider that 50 percent of American adults are not able to read an eighth-grade level book. If the FTC is to carry out MMWA's objectives, it should conduct research to investigate and require positive correlations among completed years of education and readability levels of consumers and readability levels of written materials directed to consumers. Of course, the climate in Washington since the early 1980s is not to regulate business; this view has probably kept the FTC out of this area as well as many others.

In conclusion, the evidence indicates that warranties continue to need improvement in readability. Consumers have been given a right to know the limits of the warranty and exactly what claims may be made. The suggestion is not that warranties should be liberal to the point of inviting abuse. Manufacturers also have the right to guard against unreasonable claims. However, warranties do need to be simplified for the average consumer to easily understand--it is only fair for consumers to know the terms of the warranty both before and after the sale. There may be a limited number of consumers who are motivated and have the incentive to learn about the product and terms of the warranty and, thus, do not need as much help as most consumers. In fairness to companies and their warranties, some target markets buying certain products and brands have higher educational levels than other target markets buying different products and brands. A readable warranty for the first target market might be, therefore, of a higher educational level than that for the second target market. In addition, the general climate is to keep government regulation out of business. Consequently, the FTC has no real incentive to be proactive on this issue. From a consumer standpoint, however, the FTC should require and also randomly check warranties of different products and companies to be sure warranties are being written at a level the average person can understand.


Application of Fog Index and Flesch Count to a Radio Shack Radio Warranty

The Warranty: This equipment is warranted against defects for 90 days from date of purchase. We will repair it without charge for parts and labor. Simply bring sales slip as proof of purchase date to any Radio Shack store. Warranty does not cover transport cost nor does it cover equipment subjected to misuse or accidental damage.

Number of Sentences 4

Number of Words 54

Average Sentence Length 13.5

Three or More Syllable Words 7

Hard Words per 100 13.0

Fog Index: Educational level of warranty is computed by using this formula:

Educational Level = (Average Sentence Length + Hard Words per 100) x .4 = 10.6 Years.

Flesch Count: Educational level of warranty is computed by using a weighted correction factor for personal nouns which is used in a more involved regression formula yielding the Flesch Readability Score. This score, in turn, is converted using a table into the Flesch Grade level equivalent. Because of the complexity of the computations, the results from the computer generated program are provided for illustration:

Flesch Readability Score = 64.7

Flesch Grade Level Equivalent = 9.1 Years.

In summary, the Fog Index is the easier of the two measures to apply because the average sentence length and hard words (per 100) can be determined for each warranty and the educational level is directly computed.

1 The Smog Grading Method and the Dale-Chall Method were reviewed but not used. The Smog Grading Method was not used because of questionable validity (Shuptrine and Lichtenstein 1985). Preliminary analysis of the data in this study showed inconsistencies with results from Dale-Chall, Flesch Count, and Fog Index measures. The Dale-Chall Method has college graduate only (sixteenth-grade level) as its upper scale and does not provide the range in measures on higher grade material that the authors sought.

2 A list of warranties by brand name for each product category is available from the authors.

3 For small and large durables, there is a somewhat wide range of readability levels on both measures. The range may be expected given that most of the lower readability levels were associated with dealers' brands and higher readability levels were seen among manufacturers' brands. The complexity of specific models within a product category (e.g., a 12-speed bicycle versus a single-speed bicycle) which may relate to the readability of the warranty was not examined in this research and is an excellent area for future study.


Bearden, William and Terence Shimp (1982), "The Use of Extrinsic Cues to Facilitate Product Adoption," The Journal of Marketing Research, 29(May): 229-239.

Consumer Reports (1984), "Is That Warranty Any Good?" 49(7, July): 408-410.

Federal Trade Commission (not dated), "Warranties: There Ought to be a Law ... There Is," Legal and Public Records, Washington, DC.

Flesch, Rudolf (1951), How to Test Readability, New York: Harper and Brothers.

Gunning, Robert (1952), The Technique of Clear Writing, New York: McGraw-Hill Book Company.

Kozol, Jonathan (1985), Illiterate America, Garden City, NJ: Anchor Press/Doubleday.

Lehman, John C., J. W. Gentry, L. L. Manzer, and H. W. Ellis (1982), "The Readability of Warranties: Did They Improve After the Magnuson-Moss Act and Are They More Complex Than Other Product Related Communications," unpublished working paper, Oklahoma State University, College of Business Administration, Stillwater.

MNIRAP--Minnesota Interactive Readability Approximation Program (1977), Adapted for use at the University of South Carolina by Computer Services Division Staff, Columbia.

Olson, Jerry and Jacob Jacoby (1972), "Cue Utilization in the Quality Perception Process," Proceedings: Third Annual Conference of the Association for Consumer Research, M. Venkatesan (ed.), College Park, MD: Association for Consumer Research: 167-179.

Perry, Michael and Arnon Perry (1976), "Service Contract Compared to Warranty as a Means to Reduce Consumers' Risk," Journal of Retailing, 52(2): 33-40.

Shuptrine, F. Kelly and Donald R. Lichtenstein (1985), "Measuring Readability Levels of Undergraduate Marketing Textbooks," Journal of Marketing Education, 7(Fall): 38-45.

Shuptrine, F. Kelly and Ellen Moore (1980), "Even After the Magnuson-Moss Act of 1975, Warranties Are Not Easy to Understand," The Journal of Consumer Affairs, 14 (Winter): 394-404.

U.S. Bureau of the Census (1989), Statistical Abstract of the United States 1989, 109th Edition, Washington, DC: 130-131.

Wiener, Joshua Lyle (1985), "Are Warranties Accurate Signals of Product Reliability?" Journal of Consumer Research, 12(September): 245-250.

Wilkes, Robert E. and James B. Wilcox (1976), "Consumer Perceptions of Product Warranties and Their Implications for Retail Strategy," Journal of Business Research, 4 (February): 35-42.

Wilkes, Robert E. and James B. Wilcox (1981), "Limited versus Full Warranties: The Retail Perspective," Journal of Retailing, 57(Spring): 65-77.

Wish, J. R., D. G. Steely, and S. E. Tritten (1978), The Consumer: The Art of Buying, Englewood Cliffs, NJ: Prentice Hall, Inc.: 72-76.

Ellen M. Moore is Associate Director, Graduate Studies and Interim Director, MBA Programs; and F. Kelly Shuptrine is Associate Professor, Marketing, University of South Carolina, Columbia.
COPYRIGHT 1993 American Council on Consumer Interests
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Moore, Ellen M.; Shuptrine, F. Kelly
Publication:Journal of Consumer Affairs
Date:Jun 22, 1993
Previous Article:Consumer views of the need for government intervention in the airline market.
Next Article:Cross-national comparison of consumer attitudes toward consumerism in four developing countries.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters