Help with a project, not sure where to start >o<

#1
Hi, I took a survey in which the data collected was non-numerical. For example, for most questions the options were either yes/no or often/sometimes/seldom/never, etc. For the project, I need to make inferences, but I'm not sure what to do. I tried doing regression, as I wanted to measure correlation.

What I did was assign each value a number (e.g. often was 4, never was 1). I tested this with 10 of the 115 trials, but due to the low number of possibilities there were only something like 7 points, all over the place when I did the scatterplot.

So my question is, am I on the right track, and if not, is there another, better way to make inferences based on my data?

The project is measuring, among other things, how many M-rated video games participants played (often-never), with number of real-life fights (0-5+). Any help would be greatly appreciated!
 

JohnM

TS Contributor
#2
You won't be able to make numerical inferences or estimates if the survey responses were "descriptive" (never / sometimes / etc.). Of course you could arbitrarily assign values to each response, but the validity of such assignments would be challenged.

You could treat the M-rated video games variable (often-never) as an ordinal variable, and try to show that as this variable "increases" the number of real-life fights increases (a numerical variable).