Multiple Regression and Reading Outputs with Dummy Variables?

#1
Hello!

I have another thread on here too but I hope this is significantly different to warrant another posting. So, I am using Excel and its data analysis pack. I have categorical data only and I want to explore the statistical differences in my data but I am having some trouble interpreting my outputs. I'll make sure to attach a photo for reference but I grouped my survey data into three groups from five into respondents who didn't know, respondents who said none, and respondents who said parking. My dependent variable is the binary yes or no in which respondents thought the policy was necessary or not.

When running dummy variables (as I recently re-learned) one of the 'dummies' is left out of the analysis. I chose to leave the 'didn't know' group out of it. What I am having trouble with is reading the output. It appears that there is no significance in the data set at 0.05 alpha. What does that mean? Does this output compare in any way to the people who responded 'I don't Know' ? For instance, "Respondents who said no, or parking, were not significantly different when compared to those who said I don't know" ?

Any help here is appreciated. Also, note on the screenshot -- I truncated the inputs of each field but the totals are at the bottom. For reference there were 14 'yes' responses out of 38 for the first column.

 

hlsmith

Not a robit
#2
Should you be using logistic regression given your binary dependent variable? If you had a dummy coded categorical variable, the intercept would represent the excluded variable's estimate and the other two variables would represent the expected mean increase from the left out dummy variable.
 
#3
Should you be using logistic regression given your binary dependent variable? If you had a dummy coded categorical variable, the intercept would represent the excluded variable's estimate and the other two variables would represent the expected mean increase from the left out dummy variable.
I don't know if I should be using logistic regression given the binary dependent variable but I haven't seen anything in my google searching that would suggest I couldn't... but its been a long time since I had a proper stats course. What other operation would you suggest?

--edit--
Also, for clarity, just so that I fully understand: the intercept is the missing dummy variable?
 
Last edited: