1
$\begingroup$

I am currently writing my master's thesis. To analyse the results of my survey, I conducted an ordered logistic regression using Stata. My outcome variable is whether someone wants to start a business in the future coded as followed: 1 = Definitely not, 2= Probably not, 3= Maybe, 4=Probably 5=Definitely. My first independent variable is whether someone is an Austrian coded as 0 or an immigrant coded as 1. I first ran a regression only using this IV. These are the results:

austrian0immigrant1 Odds Ratio = 2.56 p-value = 0.000

However, also want to figure out why differences between Austrian and immigrants exist therefore I tested for four entrepreneurial characteristics, which I then also included as continuous independent variables into the regression:

Austrian/Immigrant: Odds Ratio = 2.5 p-value = 0.000
Risk Tolerance Odds Ratio = 1.03 p-value = 0.458
Need for Autonomy Odds Ratio = 1.55 p-value = 0.010
Passion for Inventing Odds Ratio = 1.47 p-value = 0.001 
Need for Achievement Odds Ratio = 1.06 p-value = 0.372

I interpreted the results as followed:

The results of the ordered logistic regression prove that immigrants are more likely than Austrians to want to start their own business. The odds for immigrants to possess an intention to become an entrepreneur are 2.56 times higher than the odds for Austrians. After the inclusion of entrepreneurial characteristics into the model, the differences between Austrians and immigrants regarding the likelihood to possess an entrepreneurial intention persist. Nevertheless, instead of being 2.56 times higher, the odds for immigrants to have an entrepreneurial intention in this model are 2.5 greater than the odds for Austrians. However, as already assumed before, risk tolerance and need for achievement do not have a statistically significant influence on the intention to start a business. Nevertheless, the results of the regression prove that statistically significant relationships between entrepreneurial intention and need for autonomy and between entrepreneurial intention and passion for inventing exist. The odds of possessing an intention to become an entrepreneur are 1.55 times greater if need for autonomy increases by one unit. Likewise, if passion for inventing increases by on unit, the odds of wanting to start a business are 1.47 times greater. This indicates that need for autonomy and passion for inventing at least partly explain why immigrants are more likely to want to become an entrepreneur.

Now I wanted to make sure, whether my interpretation is right or if I made any mistakes or if there is anything relevant missing?

Thank you so much in advance.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

First, I'd avoid the word "prove".

Second, assuming you have interpreted the coding correctly (that is, that you haven't reversed Austrian and non-Austrian) then I think you are basically correct. But use the same number of decimal places throughout and the difference between 2.56 and 2.5 is probably unimportant. Finally, I'm assuming you checked the assumptions for ordinal logistic and that you sampled approrpriately and so on.

Improve this answer
$\endgroup$
7
  • $\begingroup$ If my answer satisfies you, then the usual thing is to "accept" it using the check mark. $\endgroup$
    – Peter Flom
    Commented Jul 29, 2019 at 10:47
  • $\begingroup$ Can I ask you one more thing? My research questions are: Are immigrants more likely to have an entrepreneurial intention? If yes, what are the underlying reasons behind it? I am not sure now, whether the results of the two regressions indicate that the personal characteristics at least partly explain why immigrants are more likely. So does the regression only indicate that passion for inventing and need for autonomy have a significant association with entrepreneurial intention or also that they at least partly explain why immigrants are more entrepreneurial? $\endgroup$
    – Katrin Thiem
    Commented Jul 30, 2019 at 12:09
  • $\begingroup$ The difference between the coefficients for immigrant in the two models are tiny. They don't really indicate that there was any change at all. $\endgroup$
    – Peter Flom
    Commented Jul 30, 2019 at 21:03
  • $\begingroup$ So this means that even though passion for inventing and need for autonomy have a statistically significant association with the intention the become an entrepreneur, the results do not indicate whether these two characteristics at least party explain why immigrants are more likely than Austrians to possess an entrepreneurial intention? $\endgroup$
    – Katrin Thiem
    Commented Jul 31, 2019 at 9:53
  • $\begingroup$ And I also ran another regression where I divided immigrants into from developed and from developing countries: Austrians=0 Developed=1 Developing=2. Here the OR without personal characteristics was 1.85 for developed and 4.06 for developing and after the inclusion of the personal characteristics they changed to 2.07 for developed and 3.53 for developing. So the OR increased for developed and decreased for developing. As the changes are much bigger than in the first regression, what does the increase and the decrease in the Odds Ratios tell me? Thank you in advance. $\endgroup$
    – Katrin Thiem
    Commented Jul 31, 2019 at 10:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.