Is There a Well-Defined Internal Reference Price?

ABSTRACT - This paper examines the impacts of multiple internal reference prices on consumers' evaluations of the offer. Our results show that although consumers use multiple internal reference prices in evaluating an offer, the effects of the reference prices are not mediated by the formation of a single, well-defined, unitized internal reference price (UIRP). Instead, the internal reference prices affect evaluations of the offer directly. Although the process by which internal reference prices affect the perceived offer-value is the same for the two product categories studied, internal reference price utilization is different.

Citation:

Rajesh Chandrashekaran and Harsharanjeet Jagpal (1995) ,"Is There a Well-Defined Internal Reference Price?", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 230-235.

Advances in Consumer Research Volume 22, 1995      Pages 230-235

IS THERE A WELL-DEFINED INTERNAL REFERENCE PRICE?

Rajesh Chandrashekaran, Trenton State College

Harsharanjeet Jagpal, Rutgers University

ABSTRACT -

This paper examines the impacts of multiple internal reference prices on consumers' evaluations of the offer. Our results show that although consumers use multiple internal reference prices in evaluating an offer, the effects of the reference prices are not mediated by the formation of a single, well-defined, unitized internal reference price (UIRP). Instead, the internal reference prices affect evaluations of the offer directly. Although the process by which internal reference prices affect the perceived offer-value is the same for the two product categories studied, internal reference price utilization is different.

INTRODUCTION

This paper examines the utilization of internal reference prices in evaluating the offer for two product categories. We first discuss the various definitions of internal reference price and review the literature. Then we propose two alternative models of consumer information-processing (unitization and non-unitization) and explain our data collection procedure. Finally, we discuss the empirical findings and suggest areas for future research.

DEFINING INTERNAL REFERENCE PRICE

Research on the formation and effects of internal reference prices has borrowed heavily from many psychological theories, particularly Adaptation Level Theory (Helson 1964). Considerable research supports the view that, from the consumer's perspective, price is a complex construct and is not merely the retail price (Kamen and Toman 1970; Monroe 1973; Winer 1986; Winer 1988; Jacobson and Obermiller 1990; Mayhew and Winer 1992; Rajendran and Tellis 1994). However, an extensive review of the literature reveals a wide variety of conceptual and operational definitions of internal reference price. Indeed, Winer (1988) points out that there are at least five different concepts that have been used as indicators of internal reference price.

Some studies have modeled consumers as "forward-looking." For example, Winer (1985) used "expected price" as a possible reference price for consumer durables. Jacobson and Obermiller (1990) also proposed that consumers form "expected future prices" which represent the prices consumers expect to see on their next purchase occasion which are used to decide whether to buy now or later. Based on the premise that consumers form expectations of future marketing activity from past exposure, and that these expectations are used to evaluate the current offering, Lattin and Bucklin (1989) also incorporated reference price effects in a multinomial logit model.

Other researchers view consumers as "backward-looking." A common conceptualization of reference price is as a summary of past prices (Kalwani, Kin Yim, Rinnie, and Sugita 1990; Mayhew and Winer 1992). Consistent with Adaptation Level Theory, Kalwani et al. (1990) use a weighted log-mean of the last five prices as an internal reference price. In addition to showing the reference effects of such a measure, the authors show that using a summary of past prices does not result in significantly better predictions than those obtained by simply using the price paid on the last purchase occasion.

Previous research has proposed other conceptually distinct definitions of reference price. For example, internal reference price has been studied according to consumers' perceptions of "fair price," which represents the price that consumers think is fair to charge for the product (Kamen and Toman 1970; Thaler 1985). As Emory (1970) points out, consumers' perceptions of a fair price need not correspond to any existing price. Therefore, this measure is conceptually different from a summary of past prices. Other commonly-used definitions of internal reference price include: reservation price (Scherer 1980), perceived price (Monroe 1973), recalled/evoked price (Rao and Gautschi 1982), lowest acceptable price (Stoetzel 1970), most-frequently encountered price (Olander 1970), and lowest market price (Biswas and Blair 1991).

As mentioned earlier, the literature on the definitions and effects of internal reference prices on consumer choice behavior is vast and cannot be dealt with in detail here due to space limitations. Interested readers are referred to some recent papers that provide excellent definitions and summaries of the effects of internal reference prices on consumers' purchase decisions (Klein and Oglethorpe 1987; Bearden, Kaicker, de Borrero, and Urbany 1992; Mayhew and Winer 1992; Rajendran and Tellis 1994).

Although the various definitions of internal reference price refer to constructs that are conceptually different, they all share the assumption that consumers use prices other than the retail price in relation to which the retail price is compared. As pointed out by Rajendran and Tellis (1994, p. 23), "empirical results generally support the inclusion of reference price in logit models of choice."

CRITIQUE OF THE LITERATURE

As discussed, there are numerous definitions and operationalizations of internal reference price. However, there has been no attempt to validate the different measures by incorporating them into a general model. As a result, three major issues must be addressed. First, most researchers have used single indicators of internal reference price (e.g., price last paid) and have implicitly treated these indicators as perfect measures of internal reference price. Because many research studies use internal reference price as an explanatory variable, this omission could be problematic and could lead to inconsistent parameter estimates (see Johnston 1984, pgs. 428-430). Second, previous studies have not incorporated multiple definitions of reference price in a single model to test whether the different measures capture the same underlying construct. That is, these studies have not attempted to test different theories of consumer information-processing. From a statistical viewpoint, omitting relevant price cues will lead to inconsistent parameter estimates if the omitted price cues are correlated to the included price cues (see Johnston 1984, pgs. 428-430). Third, most previous researchers have used scanner (single source) data and modeled internal reference price implicitly rather than explicitly (Mayhew and Winer 1992; Kalwani et. al 1990). As noted by Mayhew and Winer (1992, p. 64), such procedures require the strong "assumption that the specific reference price information and updating process used in the model will yield reference prices similar to the consumer's true reference prices." (Emphasis added.)

INFORMATION-PROCESSING: UNITIZATION VERSUS NON-UNITIZATION

As suggested by previous researchers (Winer 1988; Klein and Oglethorpe 1987), we acknowledge that internal reference price is a multidimensional construct. Specifically, we measure various dimensions of internal reference price and incorporate these multiple definitions into a single model which will enable us to validate the different measures and examine their relative importances in determining the final internal reference price and perceived offer-value.

FIGURE 1

UNITIZED MODEL OF REFERENCE PRICE EFFECTS

We examine the behavioral process by which consumers integrate various internal reference prices and how these internal standards affect evaluations of the offer. After reviewing the literature (Thaler 1985; Winer 1988; Klein and Oglethorpe 1987; Bearden, Kaicker, de Borrero, and Urbany 1992), we selected four commonly-used measures of internal reference price: (i) fair price, (ii) reservation price (i.e., highest price willing to pay), (iii) normal (most frequently encountered) price, and (iv) lowest price seen. Although previously paid prices have been used as proxy measures of internal reference prices, such procedures assume that consumers can recall past prices accurately. There is evidence that, in general, consumers have difficulty in recalling past prices (Dickson and Sawyer 1990). Hence, we do not use previously paid price as an indicator of internal reference price.

Kosenko and Rahtz (1988, p. 332) suggest that "consumer market price knowledge should be a concern in future price threshold studies." Rajendran and Tellis (1994) also showed that consumers use contextual prices (i.e., the prices of other brands) as reference prices. Based on these findings, we measure consumers' perceptions of "normal market price" (price most frequently encountered) as a possible internal reference price.

Having acknowledged that consumers can have multiple reference prices, we need to examine three important questions. First, are the different internal references indicators of the same underlying construct? Do consumers combine the different internal reference prices to form an overall unitized internal reference price (UIRP)? Third, what are the relative impacts of the internal prices on evaluations of the offer?

Figures 1 and 2 show two competing information-processing models by which internal reference prices can affect evaluations of the offer. The first model (see Figure 1) proposes that the different internal standards are integrated to form a single, well-defined internal reference price. We refer to this model as the "unitized" model. The second model (see Figure 2), in contrast, proposes that the various internal reference prices correspond to conceptually distinct constructs and that each affects evaluations of the offer directly. This model assumes that consumers compare incoming price cues against the internal cues independently. We refer to this information-processing strategy as the "non-unitized" model. The following section outlines the data collection procedure.

METHOD

Data were collected from 250 undergraduate student subjects. Each subject was exposed to two print ads: one for running shoes and the other for a compact disc stereo player. Consistent with retail/catalog ads in these categories, each ad contained a picture of the product, a small body of text describing key features of the product, and a retail price. The price cues were chosen to reflect the actual market prices for shoes ($34.95) and stereo players ($199) in surrounding retail stores. Approximately half the subjects first evaluated the ad for running shoes and then evaluated the ad for the stereo player, while the remaining subjects first saw the ad for the stereo player and were then exposed to the ad for running shoes. In this way, the order of presentation was counterbalanced. After viewing each ad, subjects responded to several questions pertaining to that product category.

FIGURE 2

NON-UNITIZED MODEL OF REFERENCE PRICE EFFECTS

Independent Variables

For each product category, subjects responded to the following questions that intended to measure their internal reference prices.

FAIR PRICE: I think a FAIR PRICE for the product shown would be $____.

LOWEST PRICE: The LOWEST PRICE I have seen this product for is $____.

HIGHEST PRICE: The HIGHEST PRICE I have seen this product for is $____.

NORMAL PRICE: I think the NORMAL (MOST FREQUENTLY ENCOUNTERED) PRICE for this product is $____.

Dependent Variables

The response of interest in this study is "perceived offer-value." Following Berkowitz and Walton (1980), we use three measures of offer-value:

VALUE FOR MONEY: I think the offered price represents a good value for my money.

ATTRACTIVENESS OF DEAL: I think the offered price represents a good deal.

WILLINGNESS TO BUY: I would be willing to purchase the item at the price indicated.

Each item was measured using a seven-point scale ranging from 1=disagree to 7=agree. Although all subjects responded to the same set of questions, the order of the questions was randomized across subjects to eliminate any position bias.

MODEL ESTIMATION AND EMPIRICAL RESULTS

In order to test the two information-processing models shown in figures 1 and 2 we used the structural equation methodology as implemented in LISREL 7 (J÷reskog and S÷rbom 1988). Because the two models are not nested, we needed to choose statistical criteria for comparing non-nested models. J÷reskog and S÷rbom (1993, p. 119) suggest the use of three criteria: AIC, CAIC, and ECVI, which are defined by:

AIC=^c2 + 2t

CAIC=^c2 + (1 + ln N)t

ECVI=(^c2/N) + 2(t/N)

where "t" is the number of independent parameters being estimated and "N" is the sample size. All three statistics attempt to remove the bias in favor of over-parametrized models. Specifically, the models that yield the lowest values of AIC, CAIC, and ECVI are chosen.

For each product category, we first estimated the two information-processing models (see Figures 1 and 2). In the next step, we trimmed each model by eliminating non-significant paths. Finally, we compared the two trimmed models using the three model selection criteria: AIC, CAIC, and ECVI.

Table 1 shows the results for the stereo data. The results show that the unitized model does not provide a satisfactory fit. In contrast, the non-unitized model fits the data well (^c2=6.75, d.f.=4, p=0.139) regardless of which model fit criterion is used. That is, the individual reference prices affect the perceived offer-value directly.

TABLE 1

COMPARISON OF COMPETING MODELS FOR STEREOS

FIGURE 3

NON-UNITIZED MODEL OF REFERENCE PRICE EFFECTS FOR STEREOS

The estimated parameters of the model are shown in Figure 3. The standard errors are shown in parentheses and the non-significant paths are indicated by broken lines. As expected, all parameters are positive. Only two internal reference prices (FP and HP) impact consumers' perceptions of offer-value for stereo players. These results are consistent with findings from earlier research (Kamen and Toman 1970; Thaler 1985), suggesting that consumers use some notion of "fair price" as a reference price. The use of "highest price willing to pay" (HP) lends support to the hypothesis that for such durable products as stereos, consumers use price to infer quality (within the acceptable range). In addition, the two price measures are significantly correlated (f = 0.71). Thus, using only one measure (either FP or HP) for reference price will result in inconsistent parameter estimates if internal reference price is used as an explanatory variable. Comparing the effects of the two internal reference prices on offer-value revealed no significant difference in the impacts of FP and HP on consumers' evaluations of the offer.

The results for shoes (see Table 2) are analogous to those obtained for stereos. The non-unitized model out-performs the unitized model regardless of the choice of model selection criterion (i.e., AIC, CAIC, or ECVI). Furthermore, the non-unitized model's fit is highly satisfactory (c²=7.92, d.f.=4, p=0.095). Hence, the process by which internal reference prices affect perceived offer-value is similar across the product categories studied.

The estimated model parameters are shown in Figure 4 in which the standard errors are shown in parentheses and all non-significant paths are indicated by broken lines.

TABLE 2

COMPARISON OF COMPETING MODELS FOR RUNNING SHOES

FIGURE 4

NON-UNITIZED MODEL OF REFERENCE PRICE EFFECTS FOR SHOES

FP is used as an internal reference price for running shoes. This result is similar to that obtained for stereo players. In addition, consumers also evaluate the retail price against the lowest price seen (LP). Furthermore, the two price cues are significantly correlated (f=0.58). As with the previous product category, no significant difference in the impacts of the two internal reference prices on perceived offer-value was revealed.

Overall, the results strongly support the hypotheses that consumers possess multiple cognitive reference points. Comparing the results for shoes and stereos (Figures 3 and 4) reveals that, although the process by which internal reference prices affect consumers' evaluations is the same for the two product categories, there is a major difference in the types of price cues utilized. Interestingly, fair price (FP) is used as an internal reference in both product categories. This finding has important implications: FP may serve as a reliable indicator of reference price. Thus, future studies should explicitly incorporate this measure in models dealing with the effects of reference prices on consumers' choices. In spite of this similarity in the results for both product categories, HP (which is used as an internal reference for stereos) is not used as a standard of comparison to evaluate the offer for shoes. Instead, consumers use their perceptions of the lowest market price (LP) to make judgments about offer-value. Thus, consumers' reservation prices are not always used as internal references against which retail prices are compared. These findings have important implications for how managers should convey price savings in retail ads.

DISCUSSION

In this paper, we presented some preliminary findings regarding how consumers utilize various internal reference prices and compared two information-processing models of how consumers evaluate prices. The results indicate that some IRP measures are more strongly associated with consumers' perceptions of "value" than others. Furthermore, consumers' use of internal reference prices is product specific. Consequently, it is inappropriate to use the same indicator of internal reference price for all products.

It is important to note that the results are preliminary and that the present study has some weaknesses that need to be addressed in future studies involving reference prices. First, estimates of internal reference prices were obtained after subjects were exposed to the advertised price. It is possible at least some subjects "internalized" the new information quickly and updated their internal standards. To be able to draw stronger conclusions, future studies must measure consumers' internal reference prices before and after presenting the advertised price. This will also enable researchers to understand how and when consumers accommodate new information into existing beliefs.

Future research should also replicate our study using other product categories, especially frequently-purchased products, and should allow for individual differences among consumers in level of knowledge, degree of product familiarity, and frequency of purchase.

REFERENCES

Bearden, William O., Ajit Kaicker, Melinda Smith de Borrero, and Joel E. Urbany (1992), "Examining Alternative Operational Measures of Internal Reference Prices," in John F. Sherry, Jr., and Brain Sternthal (eds.), Advances in Consumer Research, Vol. 19, Provo, UT: Association for Consumer Research, 629-635.

Biswas, Abhijit and Edward A. Blair (1991), "Contextual Effects of Reference Prices in Retail Advertisements," Journal of Marketing, 55 (July), 1-12.

Berkowitz, Eric N. and John R. Walton (1980), "Contextual Influences on Consumer Price Responses: An Experimental Analysis," Journal of Marketing Research, 17 (August), 349-358.

Dickson, P. and Alan G. Sawyer (1990), "The Price Knowledge and Search of Supermarket Shoppers," Journal of Marketing, 54 (July), 42-53.

Emory, Fred (1970), "Some Psychological Aspects of Price," in Pricing Strategy, eds. Bernard Taylor and Gordon Wills, Princeton, NJ: Brandon/Systems, 98-111.

Helson, Harry (1964), Adaptation Level Theory, New York: Harper and Row.Jacobson, Robert and Carl Obermiller (1990), "The Formation of Expected Future Price: A Reference Price for 'Forward-Looking' Consumers," Journal of Consumer Research, 16 (March), 420-432.

Johnston, Jack (1984), Econometric Methods, Mc-Graw Hill.J÷reskog, Carl, and Dag S÷rbom (1988), LISREL 7: A Guide to the Program and Applications, Chicago: SPSS, Inc.

Johnston, Jack (1993), LISREL 8: Structural Equation Modeling with the SIMPLIS Command Language, Hillsdale, NJ: Erlbaum.

Kalwani, Manohar U., Chi Kin Yim, Heikki J. Rinnie, and Yoshi Sugita (1990), "A Price Expectations Model of Customer Brand Choice," Journal of Marketing Research, 27 (August), 251-262.

Kamen, Joseph M. and Robert J. Toman (1970), "Psychophysics of Prices," Journal of Marketing Research, 7 (February), 27-35.

Klein, Noreen M.and Janet E. Oglethorpe (1987), "Cognitive Reference Points in Consumer Decision-Making," in Melanie Wallendorf and Paul Anderson (eds.), Advances in Consumer Research, Vol. 14, Provo, UT: Association for Consumer Research, 183-187.

Kosenko, Rustan and Don Rahtz (1988), "Buyer Market Price Knowledge Influence on Acceptable Price Range and Price Limits," in Advances in Consumer Research, Vol. 15, ed. Michael Houston, Provo, UT: Association for Consumer Research, 328-333.

Lattin, James M. and Randolf E. Bucklin (1989), "Reference Effects of Price and Promotion on Brand Choice Behavior," Journal of Marketing Research, 26 (August), 299-310.

Mayhew, Glenn E. and Russell S. Winer (1992), "An Empirical Analysis of Internal and External Reference Prices Using Scanner Data," Journal of Consumer Research, 19 (June), 62-70

Monroe, Kent B. (1973), "Buyers' Subjective Perceptions of Price," Journal of Marketing Research, 10 (February), 70-80.

Olander, Folke (1970), "The Influence of Price on the Consumer's Evaluation of Products and Purchases," in Bernard Taylor and Gordon Wills, eds., Pricing Strategy, Princeton, NJ: Brandon Systems Press, 50-69.

Rajendran K.N., and Gerard J. Tellis, (1994), "Contextual and Temporal Components of Reference Price," Journal of Marketing, 58 (January), 22-34.

Rao, Vithala, and David A. Gautschi (1982), "The Role of Price in Individual Utility Judgments: Development and Empirical Validation of Alternative Models," in Research in Marketing, ed. Leigh McAlister, Greenwich, CT: JAI Press.

Rosch, Eleanor (1975), "Cognitive Reference Points," Cognitive Psychology, 7 (October), 532-547.

Scherer, R. M. (1980), Industrial Market Structure and Economic Performance, Chicago, IL: Rand McNally.

Winer, Russell S. (1985), "A Price Vector Model of Demand for Consumer Durables: Preliminary developments," Marketing Science, 4 (Winter), 74-90.

Winer, Russell S. (1986), "A Reference Price Model of Brand Choice for Frequently Purchased Products," Journal of Consumer Research, 13 (September), 250-256.

Winer, Russell S. (1988), "Behavioral Perspective on Pricing: Buyers' Subjective Perceptions of Price Revisited," in Issues in Pricing: Theory and Research, ed. T. Devinney, Lexington, MA: Lexington, 35-57.

----------------------------------------