Price Intervals and Individual Price Limits As Determinants of Product Evaluation and Selection


Barbara J. Deering and Jacob Jacoby (1972) ,"Price Intervals and Individual Price Limits As Determinants of Product Evaluation and Selection", in SV - Proceedings of the Third Annual Conference of the Association for Consumer Research, eds. M. Venkatesan, Chicago, IL : Association for Consumer Research, Pages: 145-166.

Proceedings of the Third Annual Conference of the Association for Consumer Research, 1972      Pages 145-166


Barbara J. Deering, Purdue University

Jacob Jacoby, Purdue University

As research on the role of price information in consumer decisions has proliferated, an abundance of evidence for conflicting viewpoints has resulted. Attempts to evaluate the importance of price relative to other product attributes have yielded contradictory results (Andrews & Valenzi, 1971; Jacoby, Olson, & Haddock, 1971; Stafford & Enis, 1969). The relationship between price and product evaluation has been reported as both positive (Andrews & Valenzi, 1971; Gardner, 1971; McConnell, 1968; Valenzi & Andrews, 1971) and negative Lambert, 1970; Leavitt, 1954; Tull, Boring, & Gonsior, 1964). The degree to which personality traits account for the use of price information also varies across studies (Lambert, 1970; Rao, 1971).

This study attempted to clarify the function of price in product evaluation by examining objective price variations and individual differences. Variations in the objective price may have a more complex effect upon behavior than is customarily assumed. Kamen and Toman (1970) asserted that a non-Weberian relationship exists between price changes and perceived quality. Moreover, the price considered acceptable varies among individuals (Gabor & Granger, 1964, 1966) and socio-cultural groups (Sherif, 1963). These differences in latitudes of acceptance may influence the use of price information in purchase decisions.

Researchers have generally varied price by providing several explicit prices for a product. Used singly as an independent variable, this manipulation significantly alters product preference (Lambert, 1970) and judgments of perceived quality (Jacoby et al., 1971; McConnell, 1968; Valenzi & Andrews, 1971). In conjunction with other product cues, price varies in its relative importance. Andrews and Valenzi (1971) varied price, brand name, and store origin for two products. Price accounted for most of the variance in expected product quality, although all three dimensions were effective. Unfortunately, these investigators used the same subjects for all experimental conditions, and it is quite possible that demand characteristics (Orne, 1962) or other experimental artifacts confounded their data. In contrast, Jacoby et al. (1971), using different subjects for each condition, found very limited price effects when price, brand name, and product composition were varied.

Kamen and Toman (1970, p. 28) considered two price components as separate cues. At each of five absolute price levels, they varied the price of purchase alternatives by adding small amounts to the price level. This manipulation is equivalent to considering price fluctuation about a mean value. Price base refers to the mean of each range of prices considered. Price differential is the fluctuation about each mean. Base, differential, and brand name were varied for simulated effects for both base and differential on buying inclination. Interaction between base and differential occurred for one of the two brands.

Interaction between base and differential, if consistent across products, suggests that the role of price in consumer decisions depends on the particular price intervals considered. Kamen and Toman (1970) found that preference for the lower-priced of two alternatives increased with price base, but the rate of increase and differential were positively related. Nystrom (1970) reported a tendency for consumers to generalize between the price of any particular item and the generic price associated with a product assortment. Despite the apparent effectiveness of subtle price variations, researchers frequently select prices to represent approximately equal divisions of the range of market prices (Lambert, 1970; McConnell, 1968; Valenzi & Andrews, 1971).

Responsiveness to subtle price changes is basic to latitude of acceptance and rejection for price. Latitude of acceptance refers to the range of stimulus values judged acceptable by an individual, while latitude of rejection encompasses the range of objectionable stimuli (Sherif & Sherif, 1969). However, despite the fact that techniques for reliably measuring these latitudes have been described by Gabor and Granger (1966) and Monroe and Venkatesan (1969, 1970), they have not been used to clarify the function of price in consumer decisions.

A consumer should be more receptive to a price which falls within his latitude of acceptance. Because individuals differ in the particular prices that represent upper and lower limits for acceptability, price will interact with limits. An individual with a relatively low lower limit is less likely to find a price overly low than the individual with a relatively high lower limit. Similarly, a relatively low upper limit increases the likelihood that a price will be perceived as unacceptably high. Because some prices are unacceptable to almost all consumers for certain products, interaction between price and individual limits is most likely at the extremes of the product price ranges. Sherif (1963) presented prices that exceeded individuals' upper price limits. Although all subjects tended to shift their latitudes of acceptance upward, the degree of shift depended upon initial upper limits.

Because both base and differential contribute to the total price, both price components should be affected by price limits. At a price base near either limit, large rather than small differentials are more likely to produce a price which falls beyond the limit. However, the extent of differential price change is unidimensionally restricted. Theoretically, prices can increase but not decrease indefinitely. The range of price differentials which produce plausible prices appears to be more restricted near lower than near upper price limits. In this case, small differentials are more important at the lower than upper limits. Small differentials contribute more toward the possibility of an unacceptable price at low price bases. Thus, small and large differentials should evoke greater behavior change at low and high price bases, respectively. According to Kamen and Toman (1970), large differentials evoked greater change in inclination to buy regardless of base. If the price stimuli tended to fall near and beyond the upper limit, this interaction would be expected.

Given a small differential, low price base should affect behavior more than a high price base. At a high differential the reverse would be true. Kamen and Toman (1970) found that price base was more important in determining preference if the differential was large rather than small. Failure to include prices perceived as unacceptably low would also yield these results.

Without attention to price limits, interpreting the nature of relationships between price and evaluation of products is difficult. As noted, price manipulations often represent the market range. Inclusion of unacceptably low prices is unlikely. Given market prices, the lower price limit tends to approximate the lowest price stimulus presented (Gabor & Granger, 1966). Over-representation of unacceptable high prices would -elicit a negative relationship between price and perceived quality or preference without clearly detecting the relationship for acceptable prices alone. Gardner (1971) restricted prices to an acceptable range and found a positive relationship between price and likelihood of purchase. Lambert (1970) did not restrict prices ant detected a negative relationship for the same product. Although suggestive, this difference was obtained without controlling for discrepancies between experiments other than the price range used.

Two experiments were conducted to explore the simple and interactive effects of price base, differential, and latitudes of acceptance on product evaluation and selection. The first used a single measure of price limits for two products. In the second experiment, several measures of price limits were examined. The possible influence of demand characteristics on limit measures was also evaluated. Both experiments manipulated price base and differential.

Given a situation in which an individual chooses between two differently priced products which are otherwise identical, the following relationships are hypothesized. (1) Perceived quality and willingness to buy the less costly alternative are positively related to both price level and price differential. Perceived quality is frequently inferred from willingness to buy (Leavitt, 1954; Tull et al., 1964), although the two responses are clearly separable. This hypothesis assumes that price negatively affects selection of a relatively costly product, due to desire for economy, but carries connotations of quality nonetheless. (2) Price base and individual price limits interact. Perceived quality is relatively low when price falls below the lower price limit, or exceeds the upper limit. Willingness to buy is relatively low when price is unacceptably low or high and is higher for acceptable prices. (3) Price base and differential interact. Specifically, the positive relationship between perceived quality or preference for the less costly alternative and price differential is attenuated at a low price base. Small differentials tend to be more effective at a low than at a high price base, while large differentials tend toward less effectiveness at a low base. This hypothesis combines the effect of price limits for all individuals with the restriction on effective price changes at low price levels.



The subjects were 44 college students who volunteered to participate in a psychological experiment. All had purchased the products involved in the experiment at least 15 times. Twenty females and 24 males participated.

Written questionnaires were administered in a single sitting of approximately one hour. The experimenter introduced the questionnaire as a survey of consumer attitudes toward various aspects of purchase decisions. Subjects provided three basic types of information: (1) ratings of social and economic significance and quality differences for the two products under study (slacks and gasoline) as well as for 38 "filler" products; (2) highest and lowest prices considered "reasonable" for the 40 products; and (3) willingness to buy one of two alternatives, at specified prices, of slacks and gasoline.

Price limits were obtained with the question: "An average price is given for each product or service listed below. The average price was based on a sampling of stores and relevant businesses in this area. For each product, indicate the highest price that it would seem reasonable for a merchant to charge for that product. Similarly, also write the lowest price you think is reasonable. In other words, how high could the price be before you would think the product was overpriced? How low could the price be before you would find the low cost unbelievable?" Subjects wrote their responses in blanks provided. The average mean price for gasoline was $.38 a gallon, for a pair of slacks $9.40.

Price differential and base were manipulated for slacks and gasoline. The basic question for slacks was: "Suppose that you have to choose a new pair of slacks on the basis of the information given in the questions below. This is all the information you have. Which pair would you select?" Two price alternatives were presented: "Pair A sells for $ ; Pair B sells for $ ." Subjects responded on a nine-point scale containing nine equally spaced gradations labeled according to willingness to buy each of the alternatives. Endpoints were "definitely buy A" and "definitely buy B."

Nine pairs of prices, representing three differentials at each of three bases, were presented for each product. According to pretests, these price differentials assured prices beyond as well as within both upper and lower price limits. They also provided equal price intervals as manipulations. Stimulus pairs appear in Table 1.



Subjects also evaluated their chosen alternative on four 9-point semantic differential scales with endpoints of high quality-low quality, economical-uneconomical, fashionable-unfashionable, and good workmanship-poor workmanship. Perceived product quality is frequently posited as a mediator of product preference. Product attributes with financial, social, and functional implications may affect both perceived quality and preference.

Filler items were used to divide responses to price manipulations into several portions. Price manipulations were randomly selected for each portion. Order of Portions was randomized across subjects.


Measures for the five dependent variables were used in a combination of univariate and multivariate analyses of variance (Hummel & Sligo, 1971). A multivariate analysis of variance for each product examined the dependent measures as a function of price base and differential in a 3 x 3 design.

Means for highest and lowest reasonable prices were calculated for each product to represent the mean upper and lower limits, respectively. Each individual was classified according to whether each of his limits was above or below the group mean.

Following the multivariate analysis, univariate analyses of variance for each product incorporated four factors in a 3 x 3 x 2 x 2 design. Three levels of price differential and price base were within-subject factors. Classification above or below the upper and lower limits provided two levels for two between-subject factors.


Means, standard deviations, and intercorrelations of the five dependent measures appear in Table 2. Perceived workmanship, fashionableness, and quality intercorrelated highly. These three variables tended to correlate negatively with perceived economy. Willingness to buy was not significantly correlated with any other dependent measure for either product.

Given the low correlations among several of the dependent variables, a 3 (differential) x 3 (base) multivariate analysis of variance (Winer, 1971, pp. 232-240) was carried out for each product to test the hypothesis of equal population mean vectors. The hypothesis was rejected for base (F = 11.5033, df = 5,39, 2 < .001) and differential-base interaction (F = 3.9128, df = 20,558.14, 2 < .001) for slacks. Similarly, the effects of base (F = 38.0891, df = 5,39, z < .001) and interaction between differential and base (F = 5.410, df = 20,558.14, p < .001) were significant for gasoline. Base and interactive effects produced unequal treatment means for at least one dependent variable. Price differential effect was not significant for either product.

Univariate analyses of variance were conducted for each dependent variable. This procedure is not overly conservative and assures a consistent experiment-wise error rate (Hummel & Sligo, 1971). Analysis of variance for repeated measures is appropriate if two assumptions are met: homogeneity of the variance-covariance matrices and symmetry of the pooled variance-covariance matrix. In each analysis the first assumption was tested with Hartley's F max value. In no case was the hypothesis of homogeneity of variance rejected. The second assumption was not tested. Consequently, the degrees of freedom used to evaluate P values are those for the approximation procedure of Greenhouse and Geisser (1959).



Products were included as one factor in 3 x 3 x 2 analyses of variance which contained base, differential, and product effects as within-subject factors. There were no main nor interactive effects involving product differences for any dependent variable. Base, differential, and differential-base interaction produced very similar effects for four dependent measures: willingness to buy, perceived quality, fashionableness, and workmanship. Accordingly, results will be reported for only one of these variables, willingness to buy, and the fifth variable, perceived economy. Results involving perceived economy did not approximate those of any other dependent variable.

Table 3 presents the results of the 3 x 3 x 2 analysis for willingness to buy. Price base (F = 54.9057, df = 1,43, p < .001) and differential-base interaction (F = 10.0368, df = 1,43, p < .005) were significant. Inter-mean differences for price base (X = 4.3826, X2 = 3.6742, X3 = 2.6970) were significant beyond the .05 level in Newman-Keuls analysis. Willingness to buy the less costly alternative increased with price base.



Differential-base interaction appears in Figure 1. At the low price base, willingness to buy the less expensive alternative decreased with differential (X1 = 4.6818, X2 = 4.4773, X3 = 3.9886). All inter-mean comparisons for differential at this base were significant (D < .01) in Newman-Keuls tests. At the medium base, means did not differ with differential. At the high price base, means differed insignificantly for small and middle differentials. At the large differential, willingness to buy the less expensive alternative was significantly less (g < .05) than for the smaller differentials.



Results of the 3 (differential) x 3 (base) x 2 (product) analysis of variance for perceived economy appear in Table 4. Price base and differential main effects were significant (respectively, F = 12.7936, df = 1,43, E < .001; F = 20.1297, df = 1,43, g < .001). Newman-Keuls analysis indicated significantly (p < .05) greater economy at the small differential relative to either larger differential. All inter-mean comparisons for base main effects were significant (p < .01) in Newman-Keuls tests, with greatest perceived economy at the medium base (X1 = 4.1212, X2 = 5.3447, X3 = 4.5038).

Interaction between base and differential was significant (F = 27.7510, df = 1,43, p < .001). Perceived economy decreased with differential at the low base (see Figure 2) and increased with differential at medium and high bases. Mean economy did not differ in Newman-Keuls tests for middle and large differentials at the medium base. All other inter-mean comparisons were significant (g < .01).

Evaluation of price limit effects involved separate analysis of variance for each product. Each subject was classified according to whether his highest and lowest reasonable prices fell above or below the group mean for that product. For gasoline the means for high and low reasonable prices $.418 (s.d. = .0336) and $.284 (s.d. = .0519), respectively. High and low group means for slacks were $16.950 (s.d. = 4.918) and $6.617 (s.d. = 1.990). Upper and lower limits were unrelated for slacks (r = .132, n.s.) and negatively related for gasoline (r = .403, g < .01).





Analysis of variance of willingness to buy for the 2 (upper limit) x 2 (lower limit) x 3 (differential) x 3 (base) design yielded results very similar to analyses which excluded price limits. Base main effects and base-differential interaction were significant (p < .05) for both products. The pattern of interactive effects approximated that in the previous analyses. Interactions between lower limit and base for gasoline, and between upper limit and base for slacks, were marginally (p < .10) significant. For gasoline, the lower limit main effect was significant (F = 6.1226, df = 1,38, E < .05). Willingness to buy the less expensive alternative was greater for individuals with lower limits falling below the group mean.

With perceived economy of slacks alternatives as the dependent variable, analysis of variance of the 2 x 2 x 3 x 3 design yielded significant main effects for lower limit (F = 11.8576, df = 1,40, E < .005), price differential (F = 6.5176, df = 1,40, g < .001), and base-differential interaction (F = 6.4354, df = 1,40, E < .05). Perceived economy was less for individuals with a lower limit below the group mean (X = 5.4773, X = 4.3605). Newman-Keuls tests indicated significantly (p < .85) greater economy at the small differential than at middle or large differentials, which did not differ significantly. Base main effect and differential-base interaction corresponded closely to the cross-product results.

In Table 5 results of the 2 x 2 x 3 x 3 analysis of variance for perceived economy with gasoline as the product are presented. Significant base main effect (F = 4.4516, df = 1,40, z < .05) and differential-base interaction (F = 21.5931, df = 1,40, p < .001) were very similar to the cross-product results. The significant differential main effect (F = 11.2736, df = 1,40, p < .02) reflected decreases in perceived economy with increases in differential (X = 4.9156, X2 = 4.1344, X = 3.7531). All inter-mean comparisons for differential were significant (p < .01) in Newman-Keuls tests.

Three interactions among price factors and price limits were at least marginally significant: upper limit-differential (F = 3.074, df = 1,40, E < .10), both limits-differential (F = 3.1553, df - 1,40, p < .10), and lower limit-differential-base (F = 5.2694, df = 1,40, E < .05). The last three-factor interaction appears in Figure 3. In the low-lower limit condition, Newman-Keuls test indicated that perceived economy was significantly (p < .01) greater for the small differential than either of the larger differentials, which differ insignificantly. At the medium price base, economy was highest for the small differential, decreased at the large differential, and was lowest at the middle differential. All inter-mean comparisons at the medium price base were significant (p < .01). At the low base, mean economy was significantly (p < .05) greater at the middle differential than at large or small differentials.

At the low price base, within the high lower-limit condition, perceived economy was lowest for the large differential and highest for the middle differential. At the medium price base, perceived economy decreased as differential increased. Means for differentials at each base differed beyond the .01 level in Newman-Keuls analysis. At the lowest price base, mean economy differed insignificantly between middle and large differentials and decreased significantly (p < .01) at small differential.






Both price limits in Experiment 1 did not consistently interact with price manipulations as expected. It appears possible that the particular measure of price limits used may not have represented prices which were acceptable and unacceptable. Requesting prices that are "reasonable" may not elicit prices which the consumer considers acceptable for his own purchases. Consequently, a second experiment was conducted to examine several measures of price limits, including the own categories technique (Monroe & Venkatesan, 1970; Sherif & Hovland, 1961). Price base and differential were manipulated to evaluate the consistency of price effects for a third product and to examine the relationship of both price base and differential to price limits.


Written questionnaires were completed by 44 of 50 female homemakers contacted by mail and living in Indianapolis, Indiana. Potential subjects were randomly selected from a listing of residential units in the city.

Each questionnaire was prefaced with a cover letter introducing the questions about "how you make buying decisions" and emphasizing the importance of each person's ideas. Comments about the questionnaire were solicited, and participants were offered $1.00 upon receipt of their completed questionnaire.

Price base and differential were manipulated within the following basic question: "Imagine that you are in a store to buy a pair of shoes for everyday wear. You are considering two pairs that you like equally well in terms of size, style, and color. If the price of pair Y is $ __ and the price of pair Z is $ __, how likely is it that you would buy pair Y?" Four differentials ($1, $4, $7, $10) were used at each of three levels, $7.50, $19.50, and $31.50. The seven-point response scale measured willingness to buy each alternative, with endpoints of "definitely buy pair Y" and "definitely buy pair Z." Letter designations were randomized across higher and lower-priced alternatives.

Price limits were measured in four ways. One measure used an adaptation of the own categories technique (Sherif, 1963). Subjects were asked to suppose that they were in a store to purchase shoes for everyday wear. Suitable color, style, and size were assumed. A list of prices was provided to help in selecting a pair of shoes, with these instructions: "We would like for you to first divide the price list into groups of prices. To help you start, we are providing you with two category labels for these price groups. They are: TOO EXPENSIVE TO BUY (TE) and TOO CHEAP TO BUY (TC) If you find any prices that you think are too expensive to buy, write TE to the right of each price. If you find any prices you think are too cheap to buy, write TC to the right of each price. Remember, these two labels are only provided as a starting point. You do not have to use them if you find no prices that belong in these two categories."

Half of the subjects were provided with optional endpoint labels of "very low in cost" and "very high in cost." Grouping prices within brackets was demonstrated with an example. Subjects were cautioned that the number of groups or prices in a group were unimportant, as long as prices were grouped "on the basis of which prices seem to belong together." Prices ranged from $1.99 to $35.99 in $1 intervals.

Following the grouping task, subjects were asked to label one group "most acceptable" (MA), one group "most unacceptable" (MUA), and for any unlabeled groups considered acceptable (A) or unacceptable (UA), those respective labels.

Additional questions required the highest and lowest prices each subject remembered ever having paid for shoes and the highest and lowest reasonable prices. Subjects were also asked about the perceived purpose of the questionnaire, their ease in understanding the various tasks, and several demographic characteristics.


Demand Characteristics

Four indices of price limits were compared for the "too cheap/too expensive" label condition and the "too low/high in cost" label condition. The mean price, range of prices, mean lower price and mean upper price for the range of prices constituting the lower (upper) limit was computed for each set of labels. Label differences did not significantly alter any of the measures. Labeling conditions also did not affect reported price paid previously, reasonable prices, or the prices labeled acceptable and unacceptable.

Comparison of Price Limit Measures

Means and intercorrelations in Table 6 represent four measures of price limits: prices paid previously, prices considered reasonable, prices labeled very low cost (cheap) or high cost (expensive), and prices labeled acceptable or unacceptable. Previously paid and reasonable prices corresponded closely at each limit, correlating positively at both lower (r = .491, p < .001) and upper (r = .656, p < .001) limits. MUA price correlated negatively with low price paid previously (r = -.313, p < .05) or considered reasonable (r = -.470, p < .01).

The negative relationship between MUA and reasonable or previously paid prices reflects the fact that 39 of the 44 respondents classified either high or low prices, but not both, as unacceptable. The mean MUA price was $18.26, indicating that high prices tended to be unacceptable for most subjects. However, 14 subjects perceived a range of low prices as MUA. The mean MUA price for the group was $3.58. That few individuals perceived both low and high prices as unacceptable contrasts with several studies (Gabor & Granger, 1964, 1966 Sherif. 1963).

Price Limits and Price Changes

The relationship of price limits to price base and differential was examined for two indices of limits, the own categories labeling and prices considered reasonable.

Splitting the sample at mean MUA price and mean MA price provided cell entry data for a 2 x 2 x 4 x 3 analysis of variance, with two levels of each limit, four price differentials, and three price bases (Table 7). Base and differential main effects and differential-base interaction were each significant (df = 1,37, p < .05). Also significant were the MA and MUA main effects (respectively, F = 9.7611, df = 1.37, p < .005; F = 6.9725, df = 1.37, p < .05). Interaction between MA and base, and interaction among MA, differential, and base were significant beyond the .05 (df = 1.37) level. MA-differential interaction was marginally significant (p < .10, df = 1.37).

Newman-Keuls analysis indicated significant (p < .01) decrease in willingness to buy the costlier alternative between the low and medium bases (X = 3.987, X = 2.959). Mean willingness at medium and high bases differed insignificantly for differentials one and two, the lower differentials, and for the higher differentials three and four, but decreased significantly (p < .05) between the smaller and larger differentials.

Interaction between differentials and bases appears in Figure 4. At the low base, means for willingness to buy at differentials, one, two, and four differed insignificantly in Newman-Keuls tests but indicated significantly (p < .01) less tendency to buy the costlier alternative than at differential three. At the medium base, no significant differences in means occurred. Means between differentials one and two, and between three and four, differed insignificantly in Newman-Keuls analysis for differential effects at the high base. Willingness to buy the costlier alternative at the larger differentials is significantly less (p < .01) than at the smaller differentials.

Main limit effects represent significantly greater willingness to buy the costlier alternative when either the MA or MUA price is relatively high. Newman-Keuls analysis of means in the MA-differential interaction indicated significantly (p < .01) higher willingness to buy at differentials one and four for individuals with high MA prices. Mean willingness at differentials two and three did not differ according to MA limit. The significant MA-base interaction derives only from differences at the medium price base; mean willingness is greater for individuals with relatively high MA prices (p < .05 in Newman-Keuls analysis).







Although the four-factor interaction was marginally (p < .10) significant, simultaneous effects of both MA and MUA limits were of interest. Newman-Keuls tests clarified certain regularities in interactive effects. Variation in mean willingness to buy across differentials at the high price base was similar regardless of limit classification. Willingness to buy the costlier alternative tended to decrease with differential. Variation in differential affected willingness to buy at the medium base for a single group, those with relatively low MUA and MA prices. For this group, willingness to buy increased significantly from (p < .05) differentials one and two, which did not differ, to differential three, and again (p < .05) between differentials three and four. Limit effect was most notable at the low base. For low MUA price, willingness to buy the costlier alternative was greatest for differentials one and four, decreasing significantly (p < .05) at differentials two and three. For high MUA, the order of means according to decreasing willingness to buy varied with the MA limit. For high MA, the order was differential 1 = 2 = 4 > 3. For low MA, the order of decreasing willingness is differential 2 > 3 = 4 > 1.

The 2 X 2 X 4 X 3 analysis of variance with limits based on prices perceived as reasonable produced three significant effects: base (F = 25.7938, df = 1,37, p < .001); differential-base interaction (F = 6.9262, df = 1,37, p < .05); and the three-factor interaction of lower limit, differential, and base (F = 5.4706, df = 1,37, p < .05). Differential main effect, differential-lower limit interaction and the four-factor interaction were marginally (p < .10) significant. The base, differential, and differential-base effects paralleled those in the preceding analysis. Newman-Keuls analysis of means within the significant three-factor interaction revealed notable differences due to lower price limit only at the lower price base. At the medium price base, willingness to buy did not differ across differentials within each limit classification. At the high base, regardless of whether the individual had a relatively high or low lower limit, willingness to buy the costlier alternative decreased as differential increased. Decreases were significant (}! < .05) between all differentials except three and four. For the low base, given individuals with relatively low lower limits, willingness to buy was greatest at differential two and significantly less (p < .01) at any o,f the other three differentials. The lowest as well as the two highest differentials thus increased willingness to buy the less costly alternative. If lower limits were relatively high, willingness to buy the less costly alternative increased significantly (p < .05) between differential one and two, and two and three, but was lowest at the largest differential.


The positive relationship between price base and preference for less expensive alternatives predicted in Hypothesis 1 was found in both experiments, thus supporting earlier findings for low-priced, non-durable goods (Kamen & Toman, 1970; Lambert, 1970). Results of the current studies extend this relationship to two higher-priced, durable goods (slacks and shoes). Price base and perceived quality were also positively related, which confirms findings by Andrews and Valenzi (1971), Gardner (1971), McConnell (1968), and Valenzi and Andrews (1971).

Differential and willingness to buy were positively related as predicted in Hypothesis 1 for only one (shoes) of the three products studied. Because this product occurred alone in one experiment, either product-specific characteristics or experimental artifacts may be responsible for these differential effects. One explanation in terms of artifacts would involve relatively stronger differential manipulation in one experiment. Relative to mean price base, neither price differential intervals nor total range of prices within all intervals differed across products. Results may be due also to differences between samples in their attention to price. Female homemakers may be more sensitive to price differentials than students. Another possibility is increased sensitivity to differential for those who experienced four rather than three variations in differential.

Interaction between price base and differential occurred for all products, as posited in Hypothesis 3. However, the specific pattern of effects was only partially supportive. Interaction was most consistent at higher price bases. At the highest price base, preference for the less costly alternative was positively related to price differential. A higher absolute price elicited greater preference for the less costly product. At the low price base, differential usually altered dependent measures, but without any easily discernible pattern. Differential had little influence at the medium price base. Simple differential effects were consistently nonsignificant at this base. Price fluctuations about the medium base generally resulted in prices within individual price limits. Price effects appear to be relatively limited when both alternatives are clearly acceptable. Several measures of price sensitivity (Craig, Engel, & Talarzyk, 1971; Teach & Pessemier, 1966) assume this lower responsiveness for price changes within a certain range of the most acceptable price.

Price limits did not interact with price changes as predicted in Hypothesis 2. Marginally significant interactions occurred between price base and both lower limit (for gasoline) and upper limit (for slacks). The range of most acceptable prices interacted with both price level and differential for one product, shoes. At the low price base, price effect varied with price limits more than at the higher price bases. Particularly notable was the tendency of both extremely low and high differentials to produce greater preference for the more expensive alternative at the low base.- This result occurred most clearly for individuals who found low rather than high prices most unacceptable, and for those who did not think extremely low prices were reasonable. In the experiments, a large differential at the low base was most likely to produce prices below the lower limit, leaving only the costlier item as acceptable. A small differential, like a larger one, conveys higher quality connotations for the more costly alternative (Experiment 1), but is more likely to function within the limits of acceptably low prices.

Despite similar effects for price base on both willingness to buy and perceived quality, the latter two variables were unrelated in this research. Quality is often assumed to be a mediator of product preference, either as one of the import st criteria for selection or as a determinant of the salience of price (Leavitt, 1954; Tull et al., 1964). In contrast, the first experiment suggested that perceived quality and product preference constitute independent dimensions of response to price changes.

One of the more interesting findings about price base concerned s other response dimension, perceived economy, which was unrelated to willingness to buy but covaried with perceived quality. Price base and perceived economy were not related in s y simple negative manner. Perceived economy was greatest at the medium base, less at the high base, and least at the low base. Apparently, economy is not a direct reflection of financial expenditure. Evaluation of economy may encompass weighting price with associated costs over a longer period. A small present price may imply low immediate cost but higher long-term costs, perhaps through more frequent product replacement. In this case, the positive price-quality relationship may affect perceived economy.

Price differential altered perceived economy for the two products (gasoline, slacks) which were unaffected in terms of willingness to buy. Price differential and economy were negatively related. If perceived economy can be assumed to be more closely associated with price than willingness to buy, then the results suggest simply that the influence of a price change depends on how directly it is related to the dependent variable. The fluctuation in differential effect emphasizes the necessity for examining price relative to the evaluation of multiple product characteristics.

An unexpected result was the direct effect of price limits on responses. The most acceptable price was positively related to willingness to buy. Similarly, the highest reasonable price correlated positively with buying likelihood and economy, and negatively with quality. If price limits are surrogates for economic status, then the greater preference for costlier alternatives by those with greater resources is a relatively trivial finding. However, because directionality between limits and the dependent measures is not established in this study, higher acceptable prices may result from willingness or perceived economy. Numerous studies (e.g., Sherif, Sherif, & Nebergall, 1965) have shown that involvement with stimuli alters the range of stimuli considered acceptable. Willingness to buy or perceived economy, as a reflection of product importance, may affect the absolute value as well as range of prices considered. Future research should recognize the possibility of mutual influence between price limits and evaluation or selection of products.

Interaction between price limits and changes in product price varied more with differences in most acceptable prices than most unacceptable prices. Measures of most unacceptable prices provide only a gross index of an individual's emphasis on low or high prices. Most unacceptable prices, by definition, include only one portion of the price continuum, usually the upper or lower extreme. Each extreme of the continuum is discriminated more finely by price limits based on prices paid previously, considered reasonable, or perceived as "too cheap" or "too expensive." Analyses using each of these four measures indicated that each was as useful as the distinction between most unacceptable and most acceptable prices in accounting for responses to price changes. If, as this study suggests, few persons spontaneously focus on both high and low prices as unacceptable, future researchers may find it advantageous to force distinctions at both extremes.

The basic findings of these studies can be briefly summarized as follows.

(1) Preference for less costly product alternatives increased with price base. Perceived quality and price base were also positively related.

(2) Price base and differential interacted to determine response to the three products. Given a relatively high price base, preference for the less expensive alternative, as well as perceived quality, increased with price differential.

(3) Several logical definitions of price limits produced measures which did not intercorrelate highly. Demand characteristics in the experimental situation could not account for individual differences in price limits. Interaction between individual limits and price manipulations was weak and inconsistent.

(4) Price differential and perceived economy were negatively related, while perceived economy followed an inverted-V pattern relative to price base. Interaction between differential and base involved high perceived economy for large differentials at a low price base, but low perceived economy for large differentials at a high price level.

(5) Willingness to buy a costlier product alternative was unrelated to perceived quality, economy, fashionableness, and workmanship of the product.

(6) In general, the studies reveal complex interactions between price changes that are small (differential) or large (base) relative to average product price. Both base and differential price changes alter perceived economy and quality, as well as preference, for product alternatives. Changes in price base appear to be more important, in that they show more consistent relationships across products than changes in price differential. Although perceived quality and willingness to buy vary monotonically with price base, perceived economy does not.

(7) Individuals differ in their upper and lower price limits, as determined by a number of measures of limits. Limits do not apparently vary with certain demand characteristics of the experimental situation. Different measures of limits, however, are not highly related. The limits have some explanatory power, particularly when prices are relatively low. However, this efficacy for the limits is confounded with product differences and particular operational definitions for limits. Further research should concentrate on examining the possible generality of limit effects.


Andrews, I. R., & Valenzi, E. R. Combining price, brand, and store cues to form an impression of product quality. Proceedings, 79th Annual American Psychological Association, 1971, 649-650.

Craig, C. S., Engel, J. F., & Talarzyk, W. W. Consumer decision-making: On the importance of price. Proceedings, 2nd Annual Association for Consumer Research Convention, 1971, 243-255.

Gabor, A., & Granger, C. W. J. Price sensitivity of the consumer. Journal of Advertising Research, 1964, 4, 40-44.

Gabor, A., & Granger, C. W. J. Price as an indicator of quality: Report on an inquiry. Economica, 1966, 46, 43-70.

Gardner, D. M. Is there a generalized price-quality relationship? Journal of Marketing Research, 1971, 8, 241-243.

Greenhouse, S. W., & Geisser, S. On methods in the analysis of profile data. Psychometrika, 1959, 24, 95-112.

Hummel, T. J., & Sligo, J. R. Empirical comparison of univariate and multivariate analysis of variance procedures. Psychological Bulletin, 1971, 76, 49-57.

Jacoby, J., Olson, J., & Haddock, R. Price, brand name, and product composition characteristics as determinants of perceived quality. Journal of Applied Psychology, 1971, 55, 570-579.

Kamen, J. M., & Toman, R. J. Psychophysics of prices. Journal of Marketing Research, 1970, 3, 27-35.

Lambert, Z. V. Pricing and brand choice. AMA Fall Educators Conference, Boston, Massachusetts, 1970. (An abstract appears in David L. Sparks (Ed.), Broadening the concept of marketing. Chicago: American Marketing Association, 1970, p. 44.)

Leavitt, H. J. A note on some experimental findings about the meaning of price. Journal of Business, 1954, 27, 205-210.

McConnell, J. D. Effects of pricing on perception of product quality. Journal of Applied Psychology, 1968, 52, 331-334.

Monroe, K. B., & Venkatesan, M. The concept of price limits and psychophysical measurement: A laboratory experiment. In P. R. McDonald (Ed.), Marketing involvement in society and the economy. Chicago: American Marketing Association, 1969, pp.-345-351.

Monroe, K. B., & Venkatesan, M. The measurement of price thresholds: Psychophysics and latitudes of acceptance. AMA Fall Educators Conference, Boston, Massachusetts, 1970. (An abstract appears in David L. Sparks (Ed.), Broadening the concept of marketing. Chicago: American Marketing Association, 1970. p. 120.)

Nystrom, H. Retail pricing. Stockholm: The Economic Research Institute at the Stockholm School of Economics. 1970.

Orne, M. T. On the social psychology of the psychological experiment: With particular reference to demand characteristics and their implications. American Psychologist, 1962, 17, 776-783.

Rao, V. Salience of price in the perception of product quality: A multidimensional measure approach. AMA Fall Educators Conference,Minneapolis, Minnesota, 1971.

Sherif, C. W. Social categorization as a function of latitudes of acceptance and series range. Journal of Abnormal and Social Psychology , 1963, 67, 148-156.

Sherif, C. W., Sherif, M., & Nebergall, R. E. Attitude and attitude change. Philadelphia: Saunders, 1965.

Sherif, M., & Hovland, C. I. effects in communication and attitude change. New Haven: Yale University Press, 1961.

Sherif, M., & Sherif, C. W. Social psychology. New York: Harper and Row, 1969.

Stafford, J. I., & Enis, B. M. The price-quality relationship: An extension. Journal of Marketing Research, 1969, 6, 456-458.

Teach, P. D., & Pessemier, E. A. Pricing experiments, scaling consumer preferences, and predicting purchase behavior. In R. M. Hass (Ed.), Science, technology, and marketing. Chicago: American Marketing Association, 1966. pp. 541-560.

Tull, D. S., Boring, R. A., & Gonsior, M. H. A note on the relationship of price and imputed quality. Journal of Business, 1964, 37, 186-191.

Valenzi, E. R., & Andrews, I. R. Effect of price information on product quality ratings. Journal of Applied Psychology, 1971, 55, 87-91.

Winer, B. J. Statistical principles in experimental design. New York: McGraw-Hill, 1971.



Barbara J. Deering, Purdue University
Jacob Jacoby, Purdue University


SV - Proceedings of the Third Annual Conference of the Association for Consumer Research | 1972

Share Proceeding

Featured papers

See More


Time-insensitive Budget Tracking: Nudging Consumers to Spread out Spending over Time

Liang Huang, University of Arizona, USA
Anastasiya Pocheptsova Ghosh, University of Arizona, USA

Read More


F7. Mere Packaging and Consumer Choice

Tim Philipp Doering, University of Michigan, USA
Katherine Burson, University of Michigan, USA
Andrew D Gershoff, University of Texas at Austin, USA

Read More


Can Implicit Theory Influence Construal Level?

Olya Bullard, University of Winnipeg
Sara Penner, University of Manitoba, Canada
Kelley Main, University of Manitoba, Canada

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.