Selecting test - please help

Hi everybody,

I am a master student doing my first quantitative analysis.

I want to analyze if there is a relationship between a gender quota law or more specifically gender balance/imbalance on boards of directors and gender stereotypes about what it takes to be a successful board director. For those of you familiar with Schein's "Think manager - think male" theory - this is what I will try to do on the board level.

My dependent variable is: gender stereotypes about leadership (I ask the respondents to rate how much they agree that a list of 15 leadership traits is characteristic for female board directors, male board directors and successful board directors in general)
Independent variables are gender balance (ratios 50-50 or 40-60) and gender imbalance (less than 40% women) regarding board composition.

The survey is sent to Norwegian companies subjected to the quota law (at least 40% women) and companies not subjected to the quota as a control group (only those with at least 1 woman) - so this is a cross-sectional study

My hypothesis is that stereotypes are stronger in non-quota companies.

Besides, I also ask about the gender, age, board experience, educational level and background - because these could cause variances. I also ask about the share of women when respondents are from non-quota companies where I am only interested in keeping the answers of those where there is gender imbalance (less than 40% women).

The thing is, I need to know how men and women rate the traits of a successful board director and then compare it with how they rate female and male board directors. Then I need to test whether age, experience, and education cause any variance in the outcome. I guess that I have to do that for both cases (quota and non-quota companies) and then compare those as well.

What tests would you recommend?

I thank you in advance for any help and sorry if my questions are too basic. I am just a bit lost here. The more I read, the more confused I get.
Last edited:


TS Contributor
If your dependent variable can be considered as interval scaled, then a
multiple linear regression analysis, or a multilevel analysis (accounting for
clustering of respondents from the same boards, if that happened) could
be options.

With kind regards