Choosing correct statistical tests

Question

I need to come up with the fitting statistical tests for my university and don't know exactly how to approach this problem. Following scenario: In theory entrepreneurs can be divided into 2 groups: Necessity-Entrepreneurs and Opportunity-Entrepreneurs. A survey with Likert scale questions will be conducted among entrepreneurs and the data collected needs to be analysed. The 2 Hypotheses to test are:

1. Necessity entrepreneurs are less passionate about their work compared to opportunity-entrepreneurs.
2. Necessity entrepreneurs are less susceptible to cognitive bias than opportunity entrepreneurs.

Could those hypotheses tested with a simple linear regression or even simpler tests like an independent samples t-test or a X² of Independence?

Answer 1

First off, I would counsel against dichotomizing entrepreneurs between "necessity" and "opportunity". There almost certainly is a continuum between the two extremes (if this can be expressed in a single dimension at all), and dichotomizing pretends there isn't. The urge to dichotomize has been called dichotomania.

If it makes sense to capture this trait in a single dimension, and "passion" or "cognitive bias" in another dimension each, you can run standard models like correlation analyses. Since there is a priori little reason to expect a linear relationship, I would not use Pearson's correlation, but some rank-based one, like Spearman's or Kendall's.