Analysis of covariance for the removal of temporal effects in experiments with serial treatments

Will Older Models Turn Off Shoppers?

“…The use of promotional variables as treatment factors can place constraints on store selection, since stores would need to be matched on such characteristics as market potential and competition activity. Improper matching on store, area and customer characteristics could reduce the internal validity of the study (Paksoy et al, 1986).…”

Supermarket Shoppers' Response to Mature Models in Point-of-Purchase Displays

“…Three replications on the store as the test unit (Wilkinson, Mason & Paksoy 1982) were used to increase the Downloaded by [McGill University Library] at 12:52 03 November 2014 number of degrees of freedom for experimental error when only three treatments are studied. While replications of experiments increase the number of observations and the number of degrees of freedom, over an extended period of time, such as several months (Cox & Enis 1969, p. 55;Paksoy, Wilkinson & Mason 1986), replication poses the potential danger of carryover effects. The observations obtained in a given month might be distorted or biased by the effects of the experiment in previous months.…”

“…Improper matching on store, area, and customer characteristics can damage the internal validity of a field experiment (Paksoy, Wilkinson & Mason 1986). The use of one store facilitated the rotation of the POP displays and in the stock of the display bin by the researchers.…”

“…The use of one store facilitated the rotation of the POP displays and in the stock of the display bin by the researchers. Single store experiments have been successfully used in prior studies of in-store merchandising variables (Solomon, Bush & Hair, Jr. 1976;Wilkinson, Mason & Paksoy 1982;Paksoy, Wilkinson & Mason 1986).…”