Blog
Guest book

Summary and Conclusions

Overview of findings. In combination, the two studies presented in this dissertation provide partial support for the hypothesized relationships between OSL and preferred retail price environment. Study 1 confirmed all hypothesized evaluation patterns for high OSL subjects except for the one addressing differences in perceived value under "high-low" and EDLP price environments. However, low OSL subjects were found to be insensitive to price environment in their evaluation of relevant store attributes. No significant results were found in Study 2. In considering these results, three issues will be addressed: (1) the failure of Hypothesis 2, concerning differential perceived values by the two OSL groups in the two price environments, to receive any support in either of the two studies; (2) disparities in the results of the two studies; and (3) and the failure of low OSL subjects’ hypothesized differential evaluations of the two price environments to emerge. Subsequent to that, managerial implications and implications for future research will be considered.

Lack of support for the value hypothesis. Both studies failed to yield any support for the hypothesis that consumers would perceive greater values in a store featuring a price environment congruent with their respective OSL. Normally, one would expect a rational consumer, faced with a choice between two grocery stores which differ only in their pricing patterns, to choose the one that would provide him or her with the greatest value. Thus, to maintain a favorable self-perception, a consumer’s evaluation of value obtained in a store should at the very least be consistent with his or her attitudinal evaluation of the store, if not with the overall evaluation, repurchase propensity, and perceived pleasure shopping in the store. What we observe in this study, however, is that consumers do not appear to find any significant differences in values in the two stores. Thus, since neither alternative is superior to the other in terms of value provided, the consumer is free to base his or her decision on other criteria without knowingly sacrificing value. The high OSL consumer, then, can obtain value in the "high-low" store at least equal to that obtainable in the otherwise less satisfying EDLP store.

The possibility remains that some undetected relationship, possibly smaller than those represented in the positive findings for Hypothesis 1, exists between OSL and perceived value obtained in the congruent and incongruent price environments. In the present study, such effects may have failed to materialize for at least two reasons. First, it is a well known reality that large effect sizes are typically not observed in individual difference research (Kassarjian 1971; Kassarjian and Sheffet 1981). While research related to OSL appears to have done better than most other measures, as evidenced by the significant body of literature relating this variable to various consumer behavior variables, researchers’ ability to detect small or modest relationships with behavior may be limited. Secondly, the descriptions of the two stores in this study were rather terse, relying largely on statistical frequencies of their promotions or the lack hereof. This kind of description may have encouraged a focus on the explicit calculation of values obtained in the store, which, by design, were identical. In real life, in contrast, the consumer enters retail stores and is exposed to a large number of prices of different products. The consumer is not provided with any kind of summary measure of prices per se; he or she must form that impression based on his or her exposure to a number of different price exposures that must be integrated. Since the consumer rarely ever purchases exactly the same mix of goods on two shopping occasions proximate in time, a "head-to-head" bottom line is difficult to establish. Thus, had the consumer been exposed to a sequence of prices in each setting whose explicit average would have been more difficult to calculate with precision, a greater latitude for the formation of value perceptions would have existed.

Reasons for conflicting results of the two studies. The conflicting results of the two studies may be explained in part by the fact that the manipulation of stimulation inherent in the two price environments was found to be effective in the first, but not in the second, study. Further, while Study 1 employed a within-subjects design where each individual evaluated both price environments, Study 2 employed a less powerful between-subjects design.

Without doubt, the artificial setting in which consumers were tested in the laboratory experiment may in part be responsible for the lack of consistent significant results. As previously discussed, the laboratory setting of Study 2 represented a somewhat unrealistic imitation of the true shopping environment in which consumers make their purchases. In real life, the consumer faces considerably more choices, both between categories of staples and between available brands within each product category. Further, in supermarkets that offer fluctuating prices, the retailer has the ability to place large signs and/or increase the shelf space allotted to sale items, potentially creating a much more intense experience. The small space available in the setting of this experiment did not allow for any of these possibilities. The failure of this study to demonstrate a greater arousal under the "high-low" pricing condition suggests that its manipulations may have been insufficient.

A second set of problems relates to the demographics of the second study. The laboratory experiment, for logistical reasons, relied on college students. This may have introduced three sets of problems. First, college students may well have greater opportunities to engage in stimulating activities than individuals constrained by family or other responsibilities. Secondly, it became apparent when listening to many of the students that they, contrary to expectations held in planning that study, had little experience grocery shopping, living either with their parents or in an institutional setting such as university housing or a fraternal organization. Finally, subjects did not spend their own money, and, to accommodate the reality of student shopping, were asked to assume that they were shopping for only one person. The latter two limitations may have significantly reduced the stakes involved in getting a good deal.

While OSL and retail price preference may in fact be causally related, it is possible that problems in measurement and the possible presence of moderating or mediating variables, such as type of product or shopping purpose, may have obscured this relationship in the present studies. Future studies might employ several measurements of OSL as well as stronger manipulations of price environment.

Lack of support for the predicted results for low OSL subjects. Results of both studies failed to yield any support for the hypothesis that low OSL subjects would rate EDLP stores more favorably than "high-low" stores. In Study 1, where the predicted results were obtained for high OSL subjects, it was found that low OSL consumers appeared relatively insensitive to the price environment, with the differences in their evaluations not departing significantly from zero. Thus, it seems that either (1) any additional stimulation experienced in the "high-low" store did not bother the low OSL subjects or (2) additional stimulation was not experienced to a significant extent. Considering the reality that the additional stimulation felt in the "high-low" store was apparently enough to provide utility for high OSL subjects, and considering the predictive power of OSL that has been demonstrated in the large body of literature that has accumulated, the second explanation seems more plausible.

Although both high and low OSL subjects were exposed to the same descriptions, it is possible that either (1) high OSL subjects took the task more seriously and genuinely considered the merits of each environment or (2) that low OSL subjects were in some way able to ignore the stimulation inherent in the "high-low" retail environment. For the first case to be correct, one would expect that low OSL subjects would tend to rate both stores as "average" without giving much elaboration. However, as is apparent in Table 8, the sample standard deviations for the four significant variables are actually larger for the low OSL group. Thus, differences in results do not seem to be attributable to such an experimental artifact. This leaves the second explanation, that the low OSL subjects were in some way able to avoid the stimulation of the "high-low" store, as the most plausible alternative.

As mentioned in Chapter 4, it is possible that low OSL subjects might simply figure that, in the long run, the gains and losses associated with shopping for products on and off sale will roughly cancel out each other. Thus, there really would be no long term consequences of the price environment, an observation borne out by the equivalent value rating of the two formats. Consistently low OSL subjects may also be relatively low in involvement and hence less sensitive to variations in store pricing patterns. It would be interesting in future research to correlate a subject’s involvement level with his or her OSL.

Managerial implications. The present findings suggest that, at least with respect to consumer optimal stimulation level, the EDLP price environment is a dominated alternative. That is, while the retailer appears to be able to appeal better to high OSL consumers by offering a "high-low" price environment, it does not appear that a penalty is paid for this strategy in terms of appeal to low OSL consumers. This finding lends additional support for Hoch et. al.’s (1994) findings that the "high-low" format is generally more profitable for the retailer. Through their use of "margin arithmetic," Hoch. et. al. concluded that EDLP retailers could possibly maintain profitability if their price format, with lower average prices, brought about a large increase in overall volume. However, the results of this study support the assessment made by Hoch et. al.’s that the use of EDLP is unlikely to result in sales volume increases large enough to be profitable. In view of the fact that neither group rated EDLP as more favorable, and high OSL subjects tended to rate the "high-low" format more favorably, the more profitable "high-low" format does not appear to represent a competitive disadvantage.

In terms of positioning, the present study suggests that, contrary to the implications of the original hypotheses, a store’s labeling as a "high-low" store may represent little if any risk. Thus, stores that could benefit from an appeal to a concentration of high OSL consumers in their vicinity may want to make their fluctuating prices prominently visible. Since advances in information technology have made "micro" approaches to store marketing increasingly practical, it may help store managements to identify areas with predominantly high OSL consumers. Raju (1980) found evidence that OSL is positively correlated with education and negatively correlated with age and employment status, and not significantly correlated with household income. Therefore, for example, "high-low" stores may find a particularly receptive customer base in relatively younger, upwardly mobile, and higher educated urban neighborhoods. Here, the constantly changing prices may be particularly aggressively promoted through advertising of sale values.

Implications for future research. Future research is needed to explore the relationship between OSL and retail price format preference in the general population, whose demographics are more diverse than those of the more narrowly distributed subjects in the present two studies. The survey research found that the "high-low" format was preferred by a significant proportion of the population even in the Southeastern U.S., where the EDLP price format is more prevalent than in other areas (Hoch et. al. 1994). Thus, results may be even stronger in areas where the "high-low" format predominates (e.g., the Northeast; Hoch. et. al. 1994) or where the distribution is more even (e.g., California).

The mechanism by which OSL influences store evaluation deserves further research. In this study, it was found that high OSL subjects appeared to benefit from the "high-low" environment, whereas the price environment did not seem to make any difference to low OSL subjects. Since the "high-low" environment appears to provide desired stimulation for the high OSL subject, one is faced with the question of whether low OSL subjects also experience this stimulation. If they do, one would expect that this stimulation would detract from the shopping experience. Thus, future research needs to address the differential levels of stimulation experienced by the two groups of consumers in the two retail environments.

Interestingly, a great deal of variability in subject responses was found in each study. Although no clear moderator or mediator variables are evident, it is possible that a more complex relationship exists between OSL and the preferred retail price environment. For example, price environment preference may be moderated by the type of shopping task involved. The present study focused on grocery shopping. This choice of retail aimed to focus on frequently purchased products so that consumers in the "high-low" environment would be more likely to experience the fluctuating prices. Further, it was desirable to present a forum in which the motive to comparison shop is limited to avoid the ability of consumers to "cherry-pick" between competing retailers. Such a phenomenon is beyond the scope of the theoretical framework presented in this research.

Yet, under the right circumstances, shopping goods may provide a useful context for the study of consumer response to fluctuating price environments. While the motivation to do comparison shopping for higher priced items, such as consumer electronic goods, is considerably higher, the stakes for the single-store shopper are also likely to be greater. Thus, one may again find the predicted relationships among those consumers whose opportunity cost of time discourages extensive comparison shopping. In this context, it seems plausible that OSL might significantly influence the effects found by Simonson (1992) on the anticipation of regret in potentially missing out on a limited-time promotion

A separate, but highly relevant, issue for the marketer is whether any effects of optimal stimulation level are manifested directly in response to actual retail prices that the consumer faces, or more indirectly in the way the way that the retailer positions itself in terms of price stability. While consumers have been found to have relatively limited knowledge of prices that they actually pay (Dickson and Sawyer 1990), they are exposed to a great deal of advertising on store pricing. For example, while Wal-Mart, with its large share of the retail market and advertising expenditures, emphasizes stable prices, many regional retailers emphasize bargains and the acceptance of double and even triple coupons. Thus, future research might address the perceived level of price fluctuations within a given retail store rather than actual prices offered.

In Study 1, a significant contrast between the two OSL groups was found only in the case or repurchase propensity. This raises the possibility that a stronger relationship may exist between OSL and repurchase propensity than with other hypothesized predictor variables. Further research on this possibility is needed in a setting that allows for the consideration of effect sizes—e.g., causal modeling. If one can demonstrate a direct effect of OSL on repurchase propensity not mediated by other variables, an interesting observation would be made from both a theoretical and managerial perspective. From the managerial perspective, an argument can be made a store that chooses to position itself as a "high-low" retailer may want do so aggressively, in addition to merely promoting the values that are obtained. In terms of discriminant validity, it is interesting to note that the notion of hedonic shopping suggests that repurchase propensity is driven by greater perceived enjoyment in the given store setting. A significant effect within repurchase propensity may not be explainable in terms of the perceived pleasure, as measured by traditional constructs of this measure. If that turns out be the case, it may be useful to explore whether other dimensions of pleasure, enjoyment, or other hedonic benefit, may be needed to fully explain motivations for hedonic shopping.

Additional research is needed to explain the persistence of EDLP stores given the facts that (1) from an economic perspective, as demonstrated by Hoch et. al., they do not appear to maximize profitability for the retailer and (2) they do not seem to provide differential utility for either of the two segments considered in this research. Such explanations might be found either in general personality variables (e.g., attitude toward risk in particular, as opposed to general notion of OSL) or in variables more closely tied to the consumption experience (e.g., brand switching tendencies or family roles).

One interesting possibility surrounds the consumer experience in brand substitutions. When one considers the high switching rates between brands that go on and off sale (e.g., Guadagni and Little 1983), it is evident that many consumers maximize the value they obtain in "high-low" stores by disproportionately selecting the promoted brand. This raises the possibility that a consumer’s attitude toward brand switching may influence his or her evaluation of the two retail formats. One set of consumers may tend to select whatever brand is on sale in a given category and think little of the differences between the brand. A second group, traditionally termed "loyal" consumers, will tend to consistently purchase a favorite brand even when a competing one is on sale. A third group may switch brands to obtain a deal price but, unlike the first group, experience a feeling of sacrifice in giving up the preferred brand. It would be reasonable to hypothesize that the third, and possibly the second, groups would tend to value an EDLP format.

Another possible explanation for the persistence of EDLP stores may be found in the different roles that household members may serve in the shopping process. With an increase in two income households, shopping responsibilities are being shared to an increasing extent by members other than the primary homemaker. It is possible that these members have less experience shopping and are thus being directed toward the EDLP store by the homemaker who does not entrust these "amateurs" with the delicate task of optimizing value in a "high-low" store.