SPSS 1: Correlations

SPSS 1: Correlations

1. preparing for correlation analysis = creating a scatter plot

   • helps to decide which method for calcutating a correlation coefficient or whether to do one at all.
   • graphs>chart builder> scatter plot> independent on x, dependent on y

2. correlational analysis. ANALYSE> CORRELATE> BIVARIATE

   • choose Spearmans or Pearsons, one tailed or two tailed
   • MULTIPLE CORRELATION = adding all the variables into the correlation box at once.

3. partial correlation. ANALYSE > CORRELATE> PARTIAL

   • Dependent variable to the left panel. indicate which you want to control for.
     • 'when conscientiousness was held constant, there was a significant correlation between exercise and stress, r=70 =.265, p=.35

4. checking assumptions

   • Kolomogorov-Smirnov test. if MORE than 0.5 = normally distributed. ANALYZE> DESCRIPTIVES> EXPLORE. dependent factor into dependent and independent into factor
     • if not sig = in both groups the data is normally distributed.
   • Levenes = homogeneity of variance. if data is insignificant = equal variance assumed as the data is homogenous.

5. one tailed or two tailed.

   • ONE TAILED = Critical area of distribution is one sided. either greater or smaller than a certain value but not both. if it fails to fall on side hypothesis predicted; the null hypothesis will be accpeted > alternate
   • TWO TAILED = crtitical area of distribution can go in either direction. if it fails to do either; the null it taken> alternate

Imagens e arquivos do mapa