Multivariate Analysis assistance needed

#1
I have conducted a survey on online privacy sensitivity. Part of the survey used a previously-proven instrument to generate a Privacy Sensitivity Score. Other main elements of my research are for gender and for level of education.

Using ANOVA tests, I have determined that the data supports:
- Gender has no impact on the Privacy Sensitivity Score.
- As the level of formal education increases, the Privacy Sensitivity Score decreases.

My mentor said, "The statistical analysis of the impact gender and education have on the sensitivity analysis needs to be a multivariate analysis. Right now, you are using ANOVA, but it seems to me you need to relate the two variables."

Using SPSS, I have tried to do a multivariate analysis, but have not been successful. It seems to require more than 1 dependent variable. I thought the Privacy Sensitivity Score was my dependent variable and the gender and education were the fixed or covariates.

So my question, how do I conduct a multivariate analysis with the Privacy Sensitivity Score, gender, and level of education?

Thanks!
Brian
 
#5
My mentor said, "The statistical analysis of the impact gender and education have on the sensitivity analysis needs to be a multivariate analysis. Right now, you are using ANOVA, but it seems to me you need to relate the two variables."

Using SPSS, I have tried to do a multivariate analysis, but have not been successful. It seems to require more than 1 dependent variable. I thought the Privacy Sensitivity Score was my dependent variable and the gender and education were the fixed or covariates.

So my question, how do I conduct a multivariate analysis with the Privacy Sensitivity Score, gender, and level of education?

Thanks!
Brian
Hi Brian,

I think the problem is your advisor misspoke (or used a term lightly). It's one of those little word differences that I always feel picky correcting, but this is the reason why it's necessary.

As you discovered, multivariate analysis means more than one DEPENDENT variable. You need an ANOVA with two INDEPENDENT variables--gender and education. This is a two-way ANOVA, as Vinux suggested.

Do it in SPSS under General Linear Models-->Univariate. Put Gender and Education in Fixed Factors (assuming they're both categorical) and Sensitivity Score in Dependent.

Karen