from the Titanic Getting Chi-Square Statistics Data: Survival on the Titanic by Gender Analyze Descriptive Statistics Crosstabs Move one variable into ROW and the other into COLUMNS.
CLICK on CELLS Select Percentages Within Categories of the Independent Variable Independent Variable (Gender) is in the Rows Always show Observed count Optionally, show Expected count Percentage across the Rows Click CONTINUE In main dialogue box, Click STATISTICS Selecting Chi-Square Statistics Choose Chi-Square for hypothesis test Click Phi and Cramers V for measure of strength of the relationship Click CONTINUE On main dialogue box, Click OK Crosstabs Table Observed count (yellow highlight) Expected count (blue highlight) Percent within each Gender who Died or Survived (pink highlight) Report: Most men on the Titanic (80.2%) died while most women (71.6%) survived. gender * survival Crosstabulation 680.000 168.000 848.000 529.4 318.6 848.0 80.2% 19.8% 100.0% 126.000 317.000 443.000 276.6 166.4 443.0 28.4% 71.6% 100.0% 806.000 485.000 1291.000 806.0 485.0 1291.0 62.4% 37.6% 100.0% Count Expected Count % within gender Count Expected Count % within gender Count Expected Count % within gender 1 Men 2 Women gender Total 1 Died 2 Survived survival Total Chi-Square Tests 332.205 b 1 .000 330.003 1 .000 335.804 1 .000 .000 .000 331.948 1 .000 1291 Pearson Chi-Square Continuity Correction a Likelihood Ratio Fisher's Exact Test Linear-by-Linear Association N of Valid Cases Value df Asymp. Sig. (2-sided) Exact Sig. (2-sided) Exact Sig. (1-sided) Computed only f or a 2x2 table a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 166. 43. b. Chi-Square Test Pearson chi-square is the default test When Sig < alpha, variables are related. Report: The relationship is significant ( 2 (1) = 332.205, p < .005). Symmetric Measures .507 .000 .507 .000 1291 Phi Cramer's V Nominal by Nominal N of Valid Cases Value Approx. Sig. Not assuming the null hypothesis. a. Using the asymptotic standard error assuming the null hypothesis. b. Measures of Relationship Phi for 2x2 tables Cramers V for larger tables Both range from 0 to 1 with 0 = no relationship For df = 1 V = 0.10 is a small effect V = 0.30 is a medium effect V = 0.50 is a large effect Report: Gender had a large effect on chance of survival for the Titanic passengers. Survival by Social Class on the Titanic Class Social Class * Survived Died or Sur vived on Titanic Crosstabulation 117 187 304 189.8 114.2 304.0 38.5% 61.5% 100.0% 163 112 275 171.7 103.3 275.0 59.3% 40.7% 100.0% 526 186 712 444.5 267.5 712.0 73.9% 26.1% 100.0% 806 485 1291 806.0 485.0 1291.0 62.4% 37.6% 100.0% Count Expected Count % within Class Social Class Count Expected Count % within Class Social Class Count Expected Count % within Class Social Class Count Expected Count % within Class Social Class 1 Upper 2 Middle 3 Lower Class Social Class Total 0 Died 1 Survived Survived Died or Survived on Titanic Total Chi-Square Test Survival by Social Class on the Titanic Chi-Square Tests 115.246 a 2 .000 114.292 2 .000 114.311 1 .000 1291 Pearson Chi-Square Likelihood Ratio Linear-by-Linear Association N of Valid Cases Value df Asymp. Sig. (2-sided) 0 cells (.0%) have expected count less than 5. The minimum expected count is 103.31. a. Effect Size and Write-Up Symmetric Measures .299 .000 .299 .000 1291 Phi Cramer's V Nominal by Nominal N of Valid Cases Value Approx. Sig. Not assuming the null hypothesis. a. Using the asymptotic standard error assuming the null hypothesis. b. Gender was not the only factor that affected the chance of survival when the Titanic sank in the North Atlantic. Social class was significantly related to survivorship ( 2 (2) = 115.246, p < .005). More than half of the upper class passengers (61.5%) and many of the middle class (40.7%) passenger survived while only one fourth (26.1%) of the lowest class passengers survived. Social class had a moderate relationship (V = .299) to survival, not as great as the effect of gender.
For df(2): V = .07 is a small effect V = 0.21 is a medium effect V = .035 is a large effect