Univariate Categorical Measures (Frequency Analysis)

frequency analysis

counting how many cases fall into each category.

types of percentage

percent, valid percent, and cumulative

Percent

uses total sample (including missing cases)

Valid percent (most important) 

• Based only on people who actually answered 

→ this what you usually report 

Cumulative percent (only for ordered data)

• “what % of people are at this level of below"?”

• it adds categories progressively from the lowest up to the selected level.

• Cumulative percent = you are adding the previous percentages step by step down the list.

Before conducting any other type of analysis, a researcher should typically:

run frequency analyses for all variables

That’s a standard first step because it helps you detect missing data, errors, distributions, and outliers before doing more complex analysis.

univariate analysis

analysis of one variable in isolation.

A researcher reports that “177 out of 222 respondents were female.”

Which statement is most accurate?

A - This is a valid percent
B - This is a cumulative percent
C - This is raw frequency information
D - This is a confidence interval

C - This is raw frequency information

Why:

“177 out of 222” is just a count (frequency) and total sample size. It has not been converted into a percentage yet.

Multivariate analysis

when you analyze three or more variables at the same time to understand relationships among them.

bivariate analysis

analyzing relationship between two variables

Which of the following best describes why researchers run frequency analyses on all variables before other analyses?

A - To immediately test hypotheses
B - To identify missing data, errors, and unusual values
C - To compute regression coefficients
D - To calculate population parameters exactly

To identify missing data, errors, and unusual values