Year 12s: what will be the first way you research your future options? >>

SPSS nightmare

So i'm doing my dissertation and i am stuck on my spss. I'm not sure which test i should do, I did a Mann Whitney test however I don't think its right :/ I have two variables 1 being age and one being gender and the other 8 variables are candidates on a CV and with the values 1 and 2 for yes would interview and no would not interview. I wanted to see if there was a difference between which candidates are picked... someone please help i'm going mad haha

Reply 1

_Sinnie_

When you say you have 8 variables as candidates with CV values of 1 and 2, do you mean you have 8 participants who each have a value of 1 and 2?

Firstly, you're going to struggle with only 8 participants, getting sufficient power for a significant difference is unlikely.

Your dependent variable (Yes/No) is categorical, so according to my notes, you need to use a Chi Square (I think this only compares 2 variables, someone else correct me if you can compare 3 or more). This would be fine if you want to see if age has an effect and then if gender has an effect. The Wilcoxon and Mann-Whitney will only work if you're measuring a continuous variable as they work using the mean, which you don't have with 1 and 2.

Is your age continuous or categorical?

If you want to compare all three together (like an ANOVA) and age is categorical, then my notes tell me that you can either use a Cochran's Q or Logit analysis, or logistic regression (regardless of whether age is continuous or categorical). I have no idea what Logit analysis or Cochran's Q are, but as far as I know, Logistic regression would serve.

This is based on me quickly reading my notes, so don't bet your dissertation grade on it!

(edited 7 years ago)

Reply 2

acup

Hiya I went into uni today and changed how I inputted my data so instead I just did care status, values yes no and then accepted (for interview) and put next to each how many people accepted them and then did a t test worked out so much better. It got confusing but I worked it out haha thanks for replying anyway!