Statistics in SPSS Lecture 10 Petr Soukup, Charles University in Prague
CORRELATION ANALYSIS 2
Correlation coefficient Measurement of relationship between cardinal and ordinal variables Effect size for relationship Values between -1 (perfect negative) and 1 (perfect positive relationship) 0 = no relationship 3
Correlation coefficient Mostly used coefficient: Pearson’s (for cardinal), Spearman’s (for ordinal) and Kendall’s (for ordinal) Analyze-Correlate-Bivariate 4
Correlation coefficient Basic principles for individual coefficient Pearson’s (for cardinal) – covariance and standardization Spearman’s (for ordinal) – correlation for rankings Kendall’s (for ordinal) – concordance and discordance 5
Correlation coefficient Test for correlation coefficient Confidence interval for correlation 6
HW 7
HW10 Try to analyze relationship between two ordinal or cardinal variables by correlation coefficient. Interpret results. 8
Thanks for your attention 9