Q#1: The characteristics of greatest importance in predicting who will or will not purchase ETB are the characteristics 1-10 labeled "major" by Gemini. These characteristics are known to be related to buyer behavior associated with Gemini consumer products.
Q#2: The three characteristics that appear to be the most among these 10 characteristics are 1 (consecutive nonresponse to previous Gemini mailings), 3 (consecutive responses), and 7 (total dollars spent on all mail purchases). Characteristics 1 and 3 reflect buyer perceptions of Gemini consumer mail offerings, while characteristic 7 reflects the willingness of buyers to make substantial purchases through the mail.
Q#3: The best approach to the development of answers for questions 1 and 2 would be to conduct a factor analysis in which characteristics 1-10 would be assessed within the context of characteristics 11-21. The socio-economic variables would likely would be the best characteristics to use among characteristics 11-21 because four factors are explainable by 50 variables, which provides the smallest ratio between factors (dependent outcomes) and variables (independent inputs) among characteristics 11-21 when classified as "residence," "people," and "socio-demographic."
Q#4: The factors that should be considered in the selection of a similar product are price range, function, and degree of technological sophistication required of user. Each of these three factors are highly relevant to such a comparison.
Q#5: The recency of the mailing used as the basis of analysis should be considered to be crucial. Socio-demographic factors, which are highly important for these projections, can change relatively rapidly.
Q#6: The sample size of 10,000 appears to be unnecsaarialy large. With a product such as the ETB, a sample error of $1 is realistic, and a standard deviation of $5 is reasonable. Seeking a 99 percent confidence level would yield a minimum sample size re...