Tools › Statistics › Surveys and Sampling › Survey Analysis › Example
Agresti Table 3.2: gender × party identification (1991 GSS), χ² and G² tests
| Tool | Survey Analysis |
| Source | Agresti, A. (2002). Categorical Data Analysis, 2nd ed. Wiley, Hoboken, NJ. Table 3.2, Section 3.2.1 and Table 3.3. |
| Level | Undergraduate |
Problem
The 1991 General Social Survey cross-classified 980 respondents by gender and party identification: females 279 Democrat, 73 Independent, 225 Republican; males 165, 47, 191. Test independence with Pearson's X² and the likelihood-ratio G², and find which cells drive the association.
Inputs to Enter
- Default respondent rows (counted lines with a 'count' column) that reproduce the table, or the cross-tab layout: x, Democrat, Independent, Republican / Females, 279, 73, 225 / Males, 165, 47, 191
- Row variable gender, column variable party; confidence 95 %
Answers
- X² = 7.01, G² = 7.00 on 2 df, P = 0.03 (tool 7.0095 / 7.0026)
- Expected count females–Democrat 577 × 444/980 = 261.4
- Adjusted standardized residuals 2.29, 0.46, −2.62 for females (Table 3.3), +2.62 for male Republicans
- Cramér's V = √(7.0095/980) = 0.085
Notes
The tool's default data are these counts typed as respondent rows, with an illustrative satisfaction item and 0–10 recommendation score added for the Likert and NPS sections; press Submit. The same table pasted in the cross-tab layout gives identical statistics.