Chapter Fourteen. Quantitative Data Analysis
SUBGROUP COMPARISONS
"Collapsing" Response Categories
Handling "Don't Knows"
Numerical Descriptions in Qualitative Research
Moving beyond univariate analysis, it is often useful to compare subgroups. Having examined the overall frequency of attendance at religious services among the survey respondents, we might see if men or women attend more. As we'll see, this technique progresses easily into the matter of causal relationships. If we find that women earn less money, on the average, than men, is gender a "cause" of income levels?
This section will also deal with some of the mechanics of creating and reading percentage tables. Sometimes you will want to combine or collapse categories (e.g., combining "extremely liberal," "very liberal," and "somewhat liberal" into a single category called, "liberal.") We'll also look at some options for handling missing data, such as those who say "don't know" in response to an attitude question.
In Chapter 13, we saw that qualitative techniques can often reveal patterns
in quantitative data. In this section, we'll look at some of the ways
that numerical analyses can assist a basically qualitative analysis.