Although it is generally agreed that the price and perceived quality of a product are highly related, there is still a paucity of empirical evidence demonstrating the . Peterson, R. A. () The Price-Perceived Quality Relationship: Experimental Evidence. Journal of Marketing Research. 7, (November), – model that defines and relates price, perceived quality, and perceived value. Propositions about the con- cepts and their relationships are presented, then supported with evidence from experimental design and the magnitude of the price.
Affect associated with the product. While these dependent variables may in themselves be uni-dimensionaltheir collective use may show more of the dimensionality of product quality than the single measure of high and low quality. The advan- tage of these terms measures is two fold. The use of measures of product attri- butes allows researchers to determine what attributes are affected by price. Second, these measures allow the researcher to gain insight on how information is being processed by the consumer, especially how it affects an evaluation composed of several adjectives.
Overal Importance of Price While the issues raised above may result in alternative hypotheses and be partially responsible for differences in findings between studies, they leave one very important issue unresolved. The unresolved issue is the overall importance of price in determining perceived product quality. While several studies have shown that price effects perceptions of product quality 9, 19, 26, 32 other studies have shown brand name can be a more important cue than price 5, 14, 17 in determining 7 perceived product quality.
This suggests that price may not be the dominant variable, at least for all product categories. If true, only looking at the price variable may produce potentially misleading results. Specifically, several questions were advanced to guide the study. Reference group theory would tend to suggest such a finding 6 However, the reason for asking this question is that almost all product?
Does price perform the same function for products that are less conspicuous and physiological in nature? The evidence is mixed. What will be the relationship etween price and attitude toward the product? It is logical to assume that perceived product quality is really a proxy measure of attitude toward the product. If so, then we r. What will be the relationship between price and affect associated with the product? If affect is an evaluation of the product, will it be more similar to perceived quality or ratings of product attributes?
If, when talking about product quality, consumers really mean the amount of certain attributes associated with the product, then price, acting as a cue, should suggest more possession of desirable product attributes as price increases.
The above questions hint at a correlation between perceived product quality, attitude toward the product and affect associated with the product. What is the nature of that correlation? Hew important is price relative to other variables in predicting perceived product quality.
Also, tne respective levels of the independent variables ware chosen only after pretesting and ail measures of dependent variables were constructed following procedures to injure validity.
Nuisance variables were controlled by complete randomization and by incorporating into the design two major variables, brand and store. Price The exact prices used in the manipulations were established using the approach developed by Gabor and Grander 9. These two prices represent the boundary points oJ the acceptal-e price ranges. Jhe price labeled low Price Ran?
Likewise, the price labeled A bove? In all instances except two, these prices exceed the Lower and Upper Limits by 20 percent. Therefore, they should be noticeably higher or lower to a majority of subjects. The exact products to be used for this study were selected after pre-testing. From the list of products reported by Cohen and Barban, eight were selected for pre-testing. These eight products met the criterion of being of potential interest to undergraduate students.
Brand Selection The proposed design of this study called for two levels of Brand: The plan was to choose that brand as Dasirabie which was indicated by the majority of subjects as being desirable. The most undesirable brand indicated by the majority of students would be the second level of Brand, i. While there was great consistency in desirable brands, there was no consistency for un- desirable. Questioning and reflection indicated that it is impossible for an undesirable brand to remain on the market Ion-; enough for there to be agreement on its undesirabilityeven though certain brands may be undesirable to individuals.
Therefore, completely 11 unfamiliar brand names were chosen as the second level of brand. Store Selection That stores have distinct images has been shown repeatedly, starting with Martineau Therefore, it is reasonable, that for any single product, some stores are seen as a desirable place to buy a product, others as undesirable. In the pre-test, students were asked to indicate the most and least desirable store in the local area in which to purchase each of the products under consideration.
While some variance was found, it was clear that department stores and men's specialty stores were seen as the most desirable stores in which to purchase socks and shirts, with chain discount and surplus stores as the least desirable.
An almost identical pattern was found for electric toothbrush and tape recorder. The most desirable store for a tape recorder was a store special' zing in stereo systems.
Full text of "Is the price / quality relationship important ?"
The variable, store, is included in the design of this study because of the findings of Enis and Stafford 5 and Szybillo and Jacoby Enis and Stafford reported an interaction between price and store image for the product, carpeting. Szybillo and Jacoby found store image the most important variable in predicting product quality for hosiery. This indicates the possibity that store image differentially affects perceived product quality.
However, there is no data to specifically predict the exact nature of this relationship. Three-hundred-twenty-four subjects participated in the study. Each subject was exposed once to a product description of the four products. Due to random assignment of subjects to, treatment conditions, the number of subjects in each cell varied between 15 and The design of the experiment is shown in Table 1.
Each product was treated as being independent, since there was a complete randomization of treatment presentations. Dependent Variables Product quality, as measured b v a single seven-pcint scale, was the first dependent variable.
Subjects indicated their estimate of quality by placing a check on a scale having "Extremely High Quality Product" and "Extremely Low Quality Product" ds the extreme points. Willingness to Buy, as measured by a single seven-point scale, was the second dependent variable. Subjects indicated how willing they would be to buy the product if it were available at the price and store shown by placing a check on the scale having "Extremely Willing To Buy" and "Extremely Unwilling To Buy" as the extreme points on the real: Controversy surrounds any measure of attitude, especially when applied to the field of consumer behavior 3, The measure of attitude used in this study is based on a scructural model of similar algebraic form to the approaches of both Fishbein and Rosenberg 7.
The model used here has been advanced specifically for studies of consumer be- havior and is based on an expectancy x value formulation. This model advanced by Cohen and Ahtola 1 is as follows: Comparison across products, however, is severely hindered if each attribute is unique to a particular product.
Therefore, from the pre-test only those attributes were chosen for measurement that were common to three or more of the products used in the study. Durability Good construction Good materials Pleasing appearance 14 For each of these attributes, subjects first indicated to what extent the product possessed each attribu 3 by placing a che-: Subjects then indicated how important each attribute was by oiacing a check on a seven-point scale having "Not At All Important to Me" and "Very Improtant to Me" as the extreme points.
The fourth dependent variable, affect associated with the product, was composed of a series of bi-polar adjectives on seven-point scales. Through extensive factor analysis, Gardner and Ahtola 15 have developed unique sets of evaluative scales for three product categories: Once the exact product within the product category is decided upon, subjects in a pre-test can be asked to evaluate that product using the entire set of scales for the product category.
A single factor will always be the result of a factor analysis of this data. All scales loading on this single factor are then used in the study to form the measure of affect associated with the product. Therefore, the exact adjectives are unique for each product and highly loaded on the eval- uative dimension.
The number of bi-polrir adjective scales associated with each product was as follows: Manipula tions The manipulation of Brand, Store and Price was carried out by presenting subjects, in groups, packets which contained a description of a product and then the questionnaire containing the dependent variable measures. The basic product description for a product v;as identical in each treatment except for the manipulation of brand, store and price in the context of a product description.
To ensure exposure to the appropriate infomation forming the independent variables, the subject was asked to write on the questionnaire the price, store and brand name for the product being judged.
A descx typical of those used is: Exclusive 10, up and down strokes per minute. Recharges automatically when stored in stand. Comes with stand, mounting bracket and 4 brushes. Twenty minutes was allowed for comple- tion of all four questionnaires.
Products were not available for inspection. Experimental Investigation All date, was analysed using completely randomised factor: Where appropriate the Newman Keuls multiple comparison teat was used for tests of significance between means Product Quality The first dependent variable was perceived product quality. Significant differences imply that perceived product quality was affected by one of the independent variables of store, brand or pi - i': The results reported in Table 2 clearly suggest that perceived product quality is influenced by brand, store, and price, the main effects being significant in all case?
Foe each product, "desirable brand" resulted Li Lve perceived product quality than "u brand" The marginal rr. Likewise, for each product, "desirable sto're'. All possible pair-wise comparisons were significant 17 at the. For electric toothbrush - no comparisons were significant as would be indicated by the lack of a significant price main effect in Table 2. Willingness to Buy The second dependent variable was willingness to buy.
As was expected willingness to buy was generally influenced by brand, store and price. Table 4 indicates the nature of this relationship showing brand, store and price each influencing willingness to buy for the products men's socks and men's shirt. Brand and price influence willingness to buy product electric toothbrush. No significant rela- tionship was found for the product tape recorder.
Marginal means are reported in Table 5. It is clear thai: Testing for significant comparisons in willingness to buy related to levels of price the Nmrman-Keuls multiple comparison test 33 was used. For the product men's socks, all ccmparisions were significant at the.
For the products electric toothbrush and men's dress L'hirt, all comparisions were significant at the. For perspective, the data is reported in both aggregated and disaggregated form. An overall aggregated measure was constructed by multiplying, for each attribute, possession and importance scores. These products were then summed.
The F-ratios are reported in Table 6. Assuming chat an attitude toward a product is slightly different, depending on the exact composition of product characteristics and environment in a particular combination, differences should b? Table 7 contains the marginal means for the overall measure of attitude. For all products except men's socks, "desirabl-j brand" resulted in a more positive attitude than "unknown brand. These comparisons are sum- marized in Table 8.
The general relationship seems to be for attitude to be more positive as price increases. However, the two significant interactions indicate, that at least in some limited instances, store and brand may jointly influence this relationship. Items were selected which were representative of a broad range of consumer services. Further, it was desired to include items which varied with respect to respondent familiarity and the degree of quality variation within the product class.
The final list contained fifty-three services. The list of services was converted into questionnaire items of the form: Statements were then reworded into several different formats which retained the same essential meaning. For example, 'The more one pays for a servicethe better the quality," and "A service provider who charges higher prices provides better service.
A questionnaire was constructed using the fifty-three price-quality statements. In order to combat potential "ordering effects" one version of the instrument ordered the price-quality statements alphabetically by service, and a second version reversed the sequence of the four pages, renumbering items appropriately.
The two versions of the instrument were randomly distributed among the subjects. This sample is deemed appropriate given that no inferences as to proportions in the population are being sought Berkowitz and Donnerstein Study I Analyses and Results The first analysis performed was a check for order effects.
Two groups, those who received questionnaire version one and those who received version two, were compared via a univariate t test on each variable. The lack of an order effect would result in no differences between the groups; one would then expect the t-scores to be normally distributed with a zero mean. In order to reveal any underlying associations among the variables, responses to the price-quality statements were factor analyzed as per Green Nine orthogonal factors were obtained.
Based on these results, eight composite scales were formed: The scales were purified as per Churchill Resultant reliabilities and composite scale means are reported in Table 1. Further analysis compared means of the composite scales. Thus, Hypothesis 1 is supported, and one can conclude that the extent of belief in a price-quality relationship is dependent on the type of services being considered.
Based on those results, paired comparisons were performed to determine between which types of services the differences lie. The results of the paired. This was done by judgementally selecting a subset of twenty items which were believed to be representative of each major group of consumer services. Study II was designed to test research Hypotheses 2, 3, and 4.
Demographic characteristics of respondents were measured using standard methods. Variables included age measured as year of birthgender, and income total pre-tax household income. Partial correlations were used to test the hypothesis controlling for perceived variability of quality within the service category. Results are presented in Table 3. If familiarity had no effect, one would expect to see positive and negative correlations randomly distributed among the twenty service items.
Thus, it may be concluded that the data support Hypothesis 2. Partial correlations were used to test the hypothesis controlling for familiarity. Results are presented in table 3. Twelve of these were statistically significant at the. Neither of the two aberrant associations were statistically significant. Again, the findings may be interpreted as offering support for the hypothesis.
While the anticipated association does not hold true in every case, extent of belief and perceived variability cannot be said to be wholly unrelated. Positive associations were expected between extent of belief and both age and income. Partial correlations were used to test the hypotheses.
Thus, Hypothesis 4A is not supported by the data. Statistically significant associations were found in both the cases of age and income, but in neither case were the majority of the associations in the expected direction. First, the findings suggest that familiarity with a service moderates the extent of belief by reducing the role of price in forming quality perceptions.
Second, perceived variability in quality within a service category, on the other hand, may enhance the role of price as a cue to quality. Further investigation is needed to explore why perceived variability is important in some service categories but not in others.
Previous research was not affirmed by the findings relative to demographic effects. Gender differences found in previous studies might be explained by differences in product familiarity between the sexes. The list of services utilized in this study, however, was "sexually -neutral," which may account for the findings.
Perhaps one of the most interesting results of Study 11 was the unanticipated age and income effects.
The lower age group showed a tendency to believe more strongly in price-quality relationships than the older age group. Similarly, the lower income groups exhibited a greater tendency to believe in price-quality relationships than did the higher income groups. These findings contradict results of previous research in the product area. It might be argued that our results are explained by the older, more affluent subjects' greater opportunity to consume.
This may result in greater familiarity, which is negatively associated with strength of belief in price-quality relationships. However, the association was negative even when the effect of familiarity was statistically removed. An alternative explanation for the income effect may be a greater risk-aversion among lower income consumers, leading to increased use of price cues. The age effects might be alternatively explained by the limited range of ages represented in the sample.
The student sample used in the present study may be too young for direct comparison with previous literature based on adult samples. These issues are commended to future investigation. Findings indicate that the extent to which consumers believe in a price-quality relationship varies across service types, and that prior familiarity, degree of perceived quality variation within a service category, and demographic factors may influence the strength of this belief vis-a-vis certain consumer services.
However, before attempting to draw any firm conclusions from these studies, one need consider a few limitations.
In addition to the usual caveats associated with small "student samples" and self-report measures, the following limitations are noted. Second, belief in a price-quality relationship does not necessarily imply actual price reliance in the context of real purchase decisions and behaviors. Both Deering and Jacoby and Szybillo and Jacoby showed "willingness to buy" to be unrelated to "perceived quality.
Thus the findings of these studies may be generalized to pre purchase service evaluation, but not directly to behavioral intent or actual purchase. Third, one might question the representativeness of the specific services used in the studies i. Further study is needed before the results can be generalized across the entire-domain of consumer services. These limitations notwithstanding, several implications of the research may be cautiously drawn. For example, increasing a checking account service charge is not likely to result in an enhanced quality perception among prospective consumers.
However, higher-priced nursing home care may be perceived differentially as a higher quality service. The findings on familiarity effects imply that price may serve as a more important cue to service quality for 1 new services, or in 2 new markets where familiarity is limited.
This finding would-seem to favor skimming strategies over penetration pricing strategies given an appropriate service category. For example, this may hold true for new hospitality and health-related services, but is less likely for movie theaters, taxi cabs, etc. Marketers may have an opportunity, in some service industries, to attenuate the repelling attribute of price and "buy" more attracting power by altering consumers' perceptions of the extent of quality variation within a service category.
This might be done through promotional communication which emphasizes the differential qualities of various services or service providers within a category. Thus, all else being equal, consumers might be willing to pay a higher price in return for greater perceived quality. Because of the possible moderating influence of demographic variables such as age and income, markets characterized by different demographic profiles may be targeted differentially via pricing strategy.
For example, the "attracting power' of price may be stronger among younger and less affluent consumers. In addition to these managerial implications, there are several implications for future research. First, conceptual and measurement issues need to be developed further.
Second, in order to afford stronger generalizations, the matter of grouping services should be further explored. Underlying dimensions may explain differences in the magnitude of these relationships across service types. The specific nature of the differences underlying groups of services is an issue in need of further study.
Third, future research should address the untested proposition that price cues are relatively more important in the formation of service quality judgments than in tangible product quality judgments. Fourth, extent of belief in price-quality relationships among services may serve to distinguish between different groups of consumers.