The Social Norms of Discrete Consumer Exchange: Classification and Quantification - Statistical Data Included

American Journal of Economics and Sociology, The,  Oct, 1999  by Sarah Maxwell

• E-mail
• Print
• Link

To measure the magnitude of (dis)approval, the method follows Miller: it measures individuals' beliefs as to the social (dis)approval of a person's behavior within a given context and uses the mean to measure the central tendency. To determine sufficient magnitude of (dis)approval, the significant difference of the responses compared to the neutral position is determined with paired t-tests.

While the magnitude of (dis)approval is necessary, it is not sufficient for determining Decentralized Norms. The second requirement is group consensus. Extensive work has been done to develop consensus metrics, both within and between groups. It is generally agreed that unanimous consensus within a group is "virtually inconceivable" (Rossi and Berk 1985: 337); relative consensus is satisfactory. One method of measuring relative consensus is to add together the absolute value of each individual's divergence from the group mean (e.g., Spillman, Bezdek, and Spillman 1979). Another is to figure that "each individual can be different from the others by a specific constant amount... (implying that) persons agree on the norms but hold to them with varying intensity" (Rossi and Berk 1985: 339).

The purpose here, however, is not to determine consensus with an exact point on the scale but a more general consensus of perceived (dis)approval: what "most" people believe that society (dis)approves of. To measure this, the approach posited follows the use of the modal response as suggested by Jacobsen and van der Voordt (1980). To measure general (dis)approval in contrast to neutrality, it is proposed that the approval ratings be collapsed into three categories: disapproved, neutral, and approved. A criterion must then be established to determine at what point the modal proportion of a population is large enough to be able to identify a social norm. The proposed criterion for relative consensus is 65% agreement. This level allows for some variation due to lack of knowledge or understanding as well as cultural diversity, but still provides a cut-off that is clearly greater than half.

Pilot Study: A pilot study was conducted to test the ease of implementing and interpreting the results of such a scale; to determine, for example, whether respondents are capable of reporting their own beliefs as to society's (dis)approval of behaviors. The subjects were 55 undergraduate business students who participated as part of their regular class. The 40 items for the scale were developed from common pricing behaviors cited in both the literature and four focus groups. The items referred to seller's pricing behavior within the contexts of the airline and the electronic calculator industries; for example, "the seller sets the price based on competition." Respondents were asked to report the extent to which society in general (not they themselves) approved or disapproved of these behaviors.

The development of an actual scale to measure the Decentralized Norms of consumer exchange would, of course, require all the rigor of the usual tests of reliability and validity. In this pilot test, only question-order effect was tested by reversing the order of items in half the surveys (results were not significant; p [greater than] .10).