Suggestion #14 can be done in Stata using -chitest- (from SSC) as follows:
The result is:
Note the close relationship of the effect size for Chi² to Cramérs V. It would be nice if the effect size calculations could also be implemented in -chitest-.
Code:
/* Example Johnston et al. (2006, p. 414) Johnston, J.E., Berry, K.J. & Mielke Jr, P.W. (2006). Measures of effect size for chi-squared and likelihood-ratio goodness-of-fit tests. Perceptual and Motor Skills, 103(2), 412--414. */ clear input obs exp 24 8 8 10 6 10 2 12 end sum exp, meanonly local q = r(min)/r(sum) local n = r(sum) chitest obs exp di _n as txt "Effect size (Chi²) = " as res %5.4f r(chi2)/(`n'*(1-`q')/`q') _n /// as txt "Effect size (LR-Chi²) = " as res %5.4f r(chi2_lr)/(-2*`n'*ln(`q'))
Code:
. chitest obs exp observed frequencies from obs; expected frequencies from exp Pearson chi2(3) = 42.3333 Pr = 0.000 likelihood-ratio chi2(3) = 35.8661 Pr = 0.000 +-------------------------------------------+ | observed expected obs - exp Pearson | |-------------------------------------------| | 24 8.000 16.000 5.657 | | 8 10.000 -2.000 -0.632 | | 6 10.000 -4.000 -1.265 | | 2 12.000 -10.000 -2.887 | +-------------------------------------------+ . di _n as txt "Effect size (Chi²) = " as res %5.4f r(chi2)/(`n'*(1-`q')/`q') _n /// > as txt "Effect size (LR-Chi²) = " as res %5.4f r(chi2_lr)/(-2*`n'*ln(`q')) Effect size (Chi²) = 0.2646 Effect size (LR-Chi²) = 0.2786
Comment