Descriptive Statistics by Groups (Breakdown) Introductory Overview - Breakdowns vs. Frequency Tables

Another related type of analysis that cannot be directly performed with breakdowns is comparisons of frequencies of cases (n's) between groups. Specifically, often the n's in individual cells are not equal because the assignment of subjects to those groups typically results not from an experimenter's manipulation, but from subjects' pre-existing dispositions. If, in spite of the random selection of the entire sample, the n's are unequal, then it may suggest the independent variables are related. For example, crosstabulating levels of independent variables Age and Education most likely would not create groups of equal n, because education is distributed differently in different age groups. If you are interested in such comparisons, you can explore specific frequencies in the breakdowns tables, trying different orders of independent variables. However, in order to subject such differences to statistical tests, you should use the Crosstabulation Tables option which offers a variety of ways for examining crosstabulations and frequency tables, or the Log-Linear or Correspondence Analysis modules (for more advanced analyses on multi-way frequency tables).