0

I have 12 factors which affect consumer buying behavior. it is asked to the respondents(male and female) to rank these factors out of 12. 1 means most effective and 12 means least effective. I want to analyze the relationship of gender with these factors. which regression model can I apply for my analysis?

deepika kansara
  • 1

2 Answers2

1

You don't provide a lot of details, but I would think that you can just use linear regression with interaction terms. So in R, you could use 

lm(buying ~ Factor1*Gender + Factor2*Gender ... , data)

Then you will see which interactions between factors and gender are significant - those factors will then have a different importance for men and women.

Lukas Wallrich
  • 53
  • 8
  • Why linear model? The question does not specify what the nature of the dependent variable is. The question seems to also just be interested in analyzing the relationship between gender and the ranking of these factors. – AlexK May 27 '19 at 20:45
1

It appears that you have one ordinal variable (ranking of factors) and one nominal variable (gender). There is a statistical test for this kind of situation that tests whether the mean ranks are the same across your groups: a Mann-Whitney U test (also called a Mann-Whitney-Wilcoxon test or the Wilcoxon rank-sum test). There is also a more general Kruskall-Wallis test for when there are more than two groups. These are non-parametric tests that do not assume that your ranking variable is normally distributed - an assumption that underlies the two-sample t-test or the one-way ANOVA test, for example.

For more information, you can refer to these pages:

Which result to choose when Kruskal-Wallis and Mann-Whitney seem to return contradicting results?

https://www.reed.edu/psychology/stata/analyses/nonparametric/kruskal-wallacetheory.html

http://www.biostathandbook.com/kruskalwallis.html

AlexK
  • 1,007
  • 4
  • 11