Blog
Guest book

H₁: In general, high OSL subjects will respond more favorably to "high-low" outlets, whereas low OSL subjects will respond more favorably to EDLP outlets.

Specifically:

H_1a: Whereas high OSL subjects will attitudinally evaluate "high-low" outlets more favorably than EDLP outlets, low OSL subjects will evaluate EDLP outlets more favorably.

H_1b: Whereas high OSL subjects will rate the experience of shopping in a "high-low" outlet more pleasurable than shopping in an EDLP outlet, low OSL subjects will rate the EDLP shopping experience as more favorable.

H_1c: Whereas high OSL subjects will purchase more items and spend more money under "high-low" than EDLP treatments, low OSL subjects will buy more items and spend more money under the EDLP treatment.

H_1d: Whereas high OSL subjects will report a higher likelihood of returning to the store for future purchases in a "high-low" than an in EDLP outlet, low OSL subjects will report a higher likelihood of returning to an EDLP outlet.

H_1e: Whereas high OSL subjects will spend more time in a "high-low" than an EDLP outlet, low OSL subjects will spend more time in the EDLP outlet.

H₂: Whereas low OSL subjects will rate the value provided in EDLP outlets greater than that received in "high-low" outlets, high OSL subjects will rate the "high-low" outlets as providing greater value.

Although the differences in attitudes and behaviors hypothesized above would support the proposed relationship between OSL and preferred price environment, such results by themselves cannot exclude the possibility that results are driven by a spurious difference between high and low OSL consumers. It is possible that high OSL subjects may put in more effort into the shopping process, resulting in better values relative to previously held reference prices and thus justifying a greater satisfaction from transaction utility (Thaler 1985). However, since the literature cited above suggests that risk taking may, in fact, involve a net loss of resources to the individual, with the risk often being a motivating factor in and of itself, this explanation is unlikely to account for the entire result. Thus:

H₃: Differences in purchase quantities, money spent, store evaluation, perceived value, time spent in the store, and the rating of the pleasantness of the shopping experience between high and low OSL subjects cannot be accounted for completely by differences in prior reference prices and the prices of items selected for purchase.

A strong theoretical rationale has been provided for why the "high-low" retail setting should provide the high OSL subject with desired stimulation and thus lead to greater purchase behavior, more positive attitudinal evaluation, greater likelihood of returning for future purchases, and more time spent shopping in the store. However, the hypotheses above do not directly test the implicit premise that the presumed stimulation associated with the "high-low" shopping environment is responsible for any effects obtained. To test this premise, a condition that presents an alternative source of stimulation is considered. In this condition, half of the subjects were given a lottery ticket at the beginning of the experiment with instructions to wait to scratch it until the experiment had been completed, thus prolonging the suspense and arousal associated with the uncertain outcome. Under this condition, we would expect high OSL subjects confined to the EDLP retail environment to respond more favorably than they would to the simple EDLP setting. On the other hand, we would not expect as much of an effect for those high OSL subjects who would already have obtained desired stimulation from the "high-low" pricing environment. Likewise, we would not expect a significant favorable response among low OSL subjects to an alternative source of stimulation in the retail environment. Thus:

H_4a: For high OSL subjects, introducing an alternative source of stimulation into the retail environment will have a more positive effect on measures enumerated in Hypotheses 1_a through 1_e in the EDLP than in the "high-low" retail format.

H_4b: For low OSL subjects, introducing an alternative source of stimulation into the retail environment, whether EDLP or "high-low," will not have a significant positive effect on measures enumerated in Hypotheses 1_a through 1_e

The hypothesized effects are diagrammed in Figures 1 and 2.

Figure 1

Store Evaluation Under No Additional Source of Stimulation

Figure 2

Additional Source of Stimulation

In summary, it is proposed that consumers will respond more favorably to those outlets which provide a better match for their optimal level of stimulation. In terms of store pricing environments, those consumers high in need for stimulation (high OSL) will respond better to "high low" pricing, while low OSL consumers will respond more favorably to a consistent, EDLP format. This response can be manifested in terms of amounts purchased, money spent, attitudinal evaluation of the store, likelihood of returning to the store, and time spent in the store.

Method for Study 1: A Consumer Panel Study to Assess the Relationship Between Price Environment Evaluation and OSL

Research sequence

Two studies were designed to test the hypotheses proposed in this dissertation. Study 1, involving survey research based on a consumer panel, was intended to emphasize external validity. Study 2 was an experiment in a simulated store and was intended to maximize internal validity. More specifically, while Study 1 focused on attitudinal bases for a consumer’s price environment, Study 2 focused on the evaluation of actually experienced "high-low" and EDLP pricing environments. In addition, by introducing an additional source of stimulation as an added factor, Study 2 sought to rule out alternative explanations for store evaluation other than implied differences in levels of stimulation associated with each price-environment.

Some hypotheses were testable only in the context of Study 2, where subjects actually engaged in simulated shopping activity as opposed to merely evaluating the two stores described in Study 1. Specifically, Study 1 examines Hypotheses 1a, 1b, 1d, and 2. Study 2 replicates Study 1 in an experimental context and in addition tests Hypotheses 1c, 1e, 3, and 4.

Study 1

Study 1 was designed to gain insight into consumer attitudes underlying any relationship between optimal stimulation level and price environment preference. A further objective was to extend results beyond the demographically relatively homogenous frame of college students who would be available to participate in an experiment in Study 2.

A total of 550 questionnaires were sent through the Arkansas Household Panel, an omnibus panel reported to be largely representative of the state of Arkansas. This panel has been used for research published in academic marketing journals such as the Journal of Marketing, Journal of Consumer Research, Journal of Consumer Psychology, and the Journal of Retailing. 373 responses were received, for a response rate of 67.81%. It should be noted, however, that many returned questionnaires were only filled out partially, thus yielding a somewhat lower effective response rate for any given question.

The questionnaire (Appendix 2) was identified to subjects as a questionnaire on lifestyle, and participants were given instructions on the use of semantic differential scales. Immediately following was a description of either a prototypical EDLP or "high-low" store and semantic differential scales measuring attitudinal evaluation, overall evaluation, expectations of how the store would make one feel in terms of stimulation, and ratings of values received in the store. Next came the description of the contrasting store environment and evaluation items identical to those of the first store. To avoid order effects, the order of appearance of the EDLP and "high-low" stores was counterbalanced so that in half the cases, the EDLP store appeared first in the questionnaire as Store "A" while in the other half, it appeared second as store "B." Subjects were asked to imagine that the store, except for the price descriptions that followed, was "exactly like"his or her regular grocery store. The prototypical EDLP store was described as follows:

Store B never has any sales and things cost about the same every week. Store B is usually about 6% cheaper than a neighboring store on merchandise which is not on sale. However, since there are no sales in store B, you can't get anything on sale there. For example, Store B always sells a six pack of soft drinks for $2.00

The "high-low" store, in contrast, was described this way:

Store A sells about 24% of its merchandise on sale at a discount of approximately 33% off its regular prices. The items on sale vary from week to week. For example, Store A sells a six pack of soft drinks for $2.18 about three quarters of the time and puts it on sale for $1.47 about a quarter of the time.

After answering questions about the two stores, subjects were asked about the names of the stores in which they frequently shopped, how frequently they shopped in each, how much, on the average, they spent in each store, and in how many of the last ten shopping trips made for their household they had participated. Next, the descriptions of each store was repeated, and subjects were asked to make a choice the two. Finally, subjects filled out the thirty-two item OSL scale.

Measures

Although the dependent variables for this questionnaire study are conceptually identical to those employed in the laboratory study, it was necessary to employ different measurements to best accommodate consumer experiences in each setting. While the subjects in the laboratory experiment experienced an actual simulated shopping environment, survey respondents read descriptions of a store and its pricing. Thus, in this study, subjects were asked to rate the store, while subjects in the experiment rated their experience and, for some variables, their evaluations of the store based on this experience. Further, the fact that unpaid panel members evaluated two stores, as opposed to the one store evaluated by participants in the experiment, dictated that great economies in question length must be maintained. Thus, concise semantic differential scales were adopted in this study, while more wordy Likert-scale items could be used in the laboratory setting.

For each of the two prototypical price environments, subjects were asked to evaluate the store on the following measures:

Shopping pleasure. A two item scale, created by combining Swan and Futrell’s (1980) single item measure with the "pleasant-unpleasant" item from Smith and Swynyard (1983).

Attitudinal evaluation: A three item scale was created as a modified version of Smith and Swinyard’s (1983) scale. The original scale consisted of four items, but the item representing the "pleasant-unpleasant" continuum was deleted to maintain discriminant validity with the perceived pleasure variable.

Overall evaluation of the store. Overall evaluation of the store was measured on a seven point semantic differential, ranging from "Extremely low appeal" to "Extremely high appeal."

Value received for money paid. Semantic differential scales, identified by Swan and Futrell (1980), were used, as indicated in Table 5.

Perceived Arousal Associated with Store Environment (manipulation check). A seven point semantic differential scale developed by Holbrook, Chestnut, Oliva, and Greenleaf (1984), and depicted in Table 6 was employed.

Optimal Stimulation Level. Mehrabian’s (1982) Arousal Seeking Tendencies II (AST) scale was chosen from several available instruments to measure optimal stimulation level. Raju (1980) reviewed several arousal and sensation seeking tendency scales, including Zuckerman, Kolin, Price, and Zoob's (1964) Sensation Seeking Scale, Garlington and Shimota's (1964) Change Seeker Index, Penney and Reinehr's (1966) Stimulus Variation Seeking Scale, Pearson and Maddi's Similes Preference Inventory, and Mehrabian and Russell's (1974) Arousal Seeking Tendencies Scale. It should be noted that Raju clarified that the Mehrabian and Russell scale, despite its chosen name, actually measures optimal stimulation level rather than arousal seeking per se. To support this conclusion, Raju quoted from Mehrabian and Russell: "An individual's preference for an environment is closely related to his preferred arousal level; some people characteristically prefer calm settings, whereas others actively seek to increase their arousal by selecting novel, complex, or unpredictable settings." (p. 30).

Raju recommended Mehrabian and Russell's instrument for a variety of reasons: This relatively newer scale includes some of the most extensively tested and selected items from past scales, has undergone rigorous psychometric evaluation, contains fewer items than most other scales, and provides measures of arousal seeking in five separate domains (change, unusual stimuli, risk, sensuality, and new environments). A thirty-two item abbreviated version (AST-II), with superior reliability, was developed by Mehrabian (1982), but was not included in Raju's review. Wahlers, Dunn, and Etzel (1986), in testing various instruments against each other, concluded that the AST-II scale was the best measure of OSL in relation to shopping as well as risk taking and innovativeness. Steenkamp and Baumgartner (1992) concluded that the Change Seeking Index (CSI) (Garlington and Shimota (1964) provided the most consistent results across a variety of tasks. However, the AST-II provided larger effect sizes and appears to be more commonly used in the marketing literature. They thus recommend a choice between the CSI and AST-II. Given evidence from Wahlers et. al. and the more widespread use of AST-II in marketing research, this instrument seemed preferable. Although the dimensions of risk and attitude toward change may be most relevant to the present study, it was decided to employ the entire scale of thirty-two items to obtain greater reliability. The AST-II scale is only available for paid use. Due to contractual limitations forbidding the publication of actual scale items, these are not listed in this dissertation. However, they are quite similar to the original forty item AST scale. For comparison, some of the items associated with the original, published, scale include:

I feel best when I am safe and secure (reverse scored).

I don’t like the feeling of wind in my hair (reverse scored).

Sometimes I really stir up excitement.

My ideal home would be quiet and peaceful (reverse scored).

I like meeting people who give me new ideas.

I like continually changing my activities.

I avoid noisy, busy places (reverse scored).

I like to experience novelty and change in my daily routine.

I like to go somewhere different nearly every day.

I sometimes try to do things that are a little frightening.

People view me as a quite unpredictable person.

I am continually seeking new ideas and experiences.