Consumer Behavior Research Methods

Descriptive Statistics

Bivariate statistical analysis Contingency table SPSS

Always check the expected count (a) = in a 2x2 table no one should have an expected frequency less than 5, for bigger ones >20% as a rule of thumb
 
Chi Square gives you a p-value which have to be less the your significance level --> if this is the case you can be quite sure that the two variables are dependent. (Rejection of the H0 Hypothesis) --> but we don´t know the effect size yet
 
Cramers V interpretation: 
 0 = No relationship
<0.3 = Low association
0.3 - 0.6 = Medium association
>0.6 = Strong association
 

Diskussion