I am doing a correlational study and my two variables are age and scores on an emotional intelligence questionnaire. For the variable age, it is just 19-45 and there are no categories. The histograms show that both variables have a skew and there is not a normal distribution. The shapiro wilks test shows that for age it is a signifiant skew as the sig value is
Hanaa -
You first probably want to know if you have a monotonic relationship or not. I assume your data pairs the age with the "score?" Perhaps you could start by using a scatterplot to plot, say, age on the x-axis and score on the y-axis, and then you can consider where you want to go from there.
As for the skew of the two individual data sets, I don't think that is so important to what you seem to be doing. (For example, for continuous data, two skewed data sets may have a linear relationship between them ... or not. I did a lot of work in that case, where linear regression was appropriate.) I think it is important that you do that scatterplot.
[Note that "significance" is a nebulous concept without a type II error probability analysis or some other sensitivity analysis to account for effect size and sample size.]
I assume you have more than one score for the same age? That might be a reason to look at ranks. (There might be various types of regression you might want to consider though.)
I haven't done anything like this in a long time, but this looks like a good reference, and it mentions SPSS:
https://statistics.laerd.com/statistical-guides/spearmans-rank-order-correlation-statistical-guide.php
You should check at least one more source/reference as well.
Cheers - Jim
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS