Chapter 6

Chapter 6

Results of Study 2

Scale reliabilities. As indicated in Table 14, the alpha measures of internal reliability of the scales employed ranged from 0.6879 to .8883. All measures exceed 0.6, which may be acceptable for exploratory research (Nunnally 1978), which is the nature of this dissertation.

Independent variable: Optimal Stimulation Level (OSL). Subjects scored a mean value of 25.53, with a standard deviation of 33.87. These figures approximate more closely Mehrabian’s (1978) norm group means of 35 and 27, respectively, than did those for the Study 1 population. However, even the scores of the Study 2 population differ are significantly lower than those of the norm group (t₂₅₂=4.44, p<0.001). The kurtosis in this sample is 0.033, with a standard error of .306. This appears to be significantly lower than the expected value of 3.0 for a normal, symmetrically distributed population. Thus, the distribution appears to be much flatter than the normal distribution. However, cautions related to the influence of outliers on this fourth moment variable must be heeded, as discussed in Chapter 4. The skewness coefficient of -.221, with a standard error of .153, does not appear to depart significantly from zero, and thus does not suggest any major skewness in the distribution.

Manipulation checks. As part of post-experimental questionnaire, subjects answered questions about the extent of arousal they had experienced during the shopping experience. Although differences between the EDLP and "high-low" treatments were in the predicted direction, these did not reach significance (t₂₅₆ = -0.303, p=.762). Similarly, differences between those receiving and not-receiving lottery tickets went in the predicted direction, but not significantly so (t₂₅₆ = -0.362, p=0.717). Self-reported arousal did not differ significantly under the lottery and no-lottery conditions (t₂₅₄=-0.362, p=0.717).

Results. In general, the hypothesized relationships were not supported. Since the moderating effects of the presence of a lottery ticket as an additional source of stimulation did not come about, a set of tables collapsed across this factor is presented in addition to that involving the complete factorial. Mean values of the dependent variables for the full factorial are indicated by condition in Table 15, while Table 16 contains values for the reduced 2x2 OSL x price-environment factorial relevant to most hypotheses.

Table 15 Means of Dependent Variables by Condition—Full Factorial
	Condition
Variable	Low OSL				High OSL



Attitude	10.14	11.62	11.58	11.16	9.61	11.46	11.04	9.92
	(3.69)	(3.83)	(4.75)	(3.39)	(3.76)	(3.83)	(4.81)	(4.67)
	35	27	33	25	36	24	41	27

Repurchase	9.03	9.26	10.63	8.68	8.05	10.50	9.14	8.78
Propensity	(3.35)	(3.83)	(4.89)	(3.71)	(3.55)	(3.60)	(4.85)	(4.19)
	35	27	33	25	36	24	41	27

Value	14.03	14.89	15.88	13.83	13.80	15.00	15.78	14.67
Received	(3.35)	(5.56)	(6.43)	(3.47)	(3.62)	(3.80)	(6.88)	(5.15)
	35	27	33	24	36	24	41	27

Money	37.99	33.18	31.89	31.93	36.37	36.03	36.47	34.78
spent	(10.99)	(11.11)	(11.55)	(12.84)	(11.24)	(12.92)	(12.51)	(11.09)
	36	27	33	26	36	24	39	27

Time	4.23	3.67	3.55	4.22	5.19	3.78	4.21	4.21
spent in	(2.25)	(1.91)	(1.66)	(2.96)	(6.22)	(1.80)	(2.28)	(1.69)
store	36	26	33	26	36	24	41	27

No. of	21.92	18.70	19.36	18.84	21.91	21.58	21.52	20.56
items	(6.60)	(6.41)	(6.67)	(6.90)	(7.40)	(6.90)	(7.69)	(6.61)
purchased	36	27	33	26	36	24	39	27

Hypothesis 1 suggested that high OSL subjects would rate the "high-low" price environment more favorably than the EDLP environment and that this pattern would be reversed for low OSL subjects. Specifically, Hypothesis 1a considered attitudinal evaluation, where no significant differences were found (F_1,238=0.668, p=.415).

Hypothesis 1b proposed that a congruent price-environment would lead to the purchase of more items and the spending of more money. Results indicate that there is no statistical interaction with respect to items purchased (F_1,240=0.280, p=0.597), nor with the total amount spent in the store (F_1,238=2.297, p=0.131).

Hypothesis 1c suggested a greater likelihood of return to a store featuring a congruent price environment; however, differences are not significant (F_1,238=0.227, p=0.634). Interestingly, although not directly hypothesized, a significant interaction of OSL and the presence of a lottery ticket emerges (F_1,238=3.955, p=0.048). In view of the failure of the manipulation check, however, this interaction does not appear to be meaningfully interpretable.

Hypothesis 1d proposed that an individual would spend more time in a price/OSL congruent environment. Results indicate no significant difference (F_1,242=1.109, p=0.293).

Hypothesis 2 suggested that individuals would perceive a greater value received in a store congruent with their optimal stimulation level. This hypothesis was not supported (F_1,238=0.052, p=0.821).

Hypothesis 3 proposed that the previously explained relationships could not be explained in terms of the actual prices of the items received relative to the prices that would have prevailed in a contrasting condition. In view of the absence of significant results in the prior hypotheses, Hypothesis 3 is not testable.

Hypothesis 4 held that an alternative source of stimulation, a lottery ticket, could serve to raise the stimulation level in the EDLP condition and achieve a favorable effect comparable to that of the "high-low" pricing environment among high OSL subjects. Since price-environment effects were not detected, this hypothesis is not testable.

Discussion. The lack of significant results in this study are consistent with the apparent failure of the manipulations to achieve the desired results. Although the "high-low" retail price-environment in this study was found to be no more arousing nor risky than that of EDLP, this finding may be attributable to limitations in this study. The current simulated store contained only fifty-two products, and was confined to a 10 by 4.5 feet room, factors which may have impaired the realism of the lab.

A further limitation of the present laboratory setting, relative to an actual supermarket environment, is that the latter often provides opportunities for much more extensive displays of sales than was possible in the laboratory. While it is possible, in the field, to devote extensive displays and signage to sale items, this was not possible in the tightly confined laboratory environment. Under actual shopping conditions, subjects further take possession of the items they have purchased, which may add significantly to the experience.

It should be noted that experimental conditions may take out much of the richness of the shopping situation available in a large supermarket featuring a large selection. In that setting, the subject faces the choice between selecting his or her favorite brand, while not on sale, or settling for a less preferred brand currently featured on deal.

Other problems have become apparent in the choice of a student population. First, many subjects indicated that they did not actually do grocery shopping, living either at home with families or relying on food provided by the university food services. Secondly, it should be noted that students, when they shop, frequently shop for only one person. Thus, the "stakes" on the shopping outcome, in terms of money saved or lost, significantly smaller than they would for someone shopping for an entire family. Finally, it should be noted that students, typically not having family responsibilities and, being younger, are better able to pursue other forms of stimulating activities, and thus may not find shopping to be as useful an outlet for stimulation seeking than the general population would. Also, although students are generally a relatively homogenous group, it should also be noted that the experiment was conducted at a university with a very large foreign-born population, whose shopping habits may differ from those of the traditional American shopper. For example, the limited assortment of goods in the store might have made it difficult to compose a balanced vegetarian diet.

An additional potential problem involves the reality that shopping in a simulated store for the first time is inherently a novel situation, thus resulting in a stimulating experience. Although this problem might be overcome by having each participant come to the lab several times, it would not have been feasible under the present circumstances.