Correlation between scale vars controlling for nominal and ordinal vars - Which test?

#1
Hello all,

I am not completely sure if the SPSS forum is the correct place to post this, so aplogises in advance if it's not. However, I am using SPSS to perform statistical analysis on my data, so at some point I will have to implement whichever test is needed in SPSS, hence my choice of forum.

I have a data set consisting of about 20 patients. The variables include age (I use an ordinal age group variable, 1, 2 or 3), gender (nominal, 0 or 1), time since the patient was diagnosed (months, scale), and couple of hundred variables which are quantitative scale data extracted from MRI-images corresponding to different regions of interest.

What I want to do is to check for correlation between each of these MRI-variables (separately) and the time since diagnosis, while controlling for age and gender.

I was going to use the partial correlation function, but then I noticed here (http://www.zu.ac.ae/main/files/contents/research/training/intermediate_pasw.pdf) that it is only to be used if all variables are scale variables. I was thinking about using cross-tabs, as suggested in the same reference, with the control variables as layers, but was unsure what the difference between the chi-square test and the correlation test are, and which is more applicable to my problem. Reading the SPSS documentation unfortunately didn't make my choice easier, as I am, although not a complete newbie, still relatively new to statistics.

Could someone please help explain to me how best to proceed with this problem. Thank you in advance!