Proposal design-is a control necessary?

I have to design a proposal in order to evaluate 'cross-racial' attitudes of young people who live in a high crime area,towards gun carrying. Can I simply evaluate attitudes, and compare them across different races, or do I need to have young people who live in a low crime area to act as a comparison or a control? Would this help me determine if it's the race or the area that effects the attitude? Also, which test would I use in either of these scenarios, I'm thinking a t-test or ANOVA? Thanks!


TS Contributor
It would help you evaluate the interaction between race and area. In other words, is the difference in race attitudes modified by the area?

So, I think it would be a good idea....

....and you would use ANOVA.
Now what?!!!

Hi John,

Thanks for replying to my last thread! Now I am confused again-I have completed my proposal, and as suggested have included participants from a low crime area as well. I have decided to evaluate attitudes using a semi-structured interview technique. However, as this will be produce qualitative data can I use ANOVA or will I have to use a package such as QSR N5? Is there any way I can transform the responses/attitudes into quantitative data?

P.S. The interview will consist of open and closed questions.
Last edited:


TS Contributor

The open-ended responses are pretty tough to deal with - maybe the frequency of "key words"?

The closed questions - if they're multiple-choice, then it's obvious.