rydingray
06-21-2006, 09:19 AM
Hi,
I am performing chi-square analyses for my dissertation data and am wondering about appropriate "post-hoc" analyses when chi-square values are significant. I read one resource (Sheskin, 2004) that says to look for standardized residuals whose absolute values are greater than 1.96 but another resource (Haberman, 1973) says adjusted residuals are the data to examine. I like the idea of using adjusted because several of them are greater than 1.96 and thus seem to identify sources of chi-square significance whereas all the standardized values are less than 1.96 and thus don't seem to identify sources of chi-square significance. Anyone know for sure which if these, if either, are the appropriate post-hoc data to examine? Thanks!
Brian.
Haberman, S.J. (1973). The analysis of residuals in cross-classified tables. Biometrics, 29, p. 205-220.
Sheskin, D.J. (2004). Handbook of Parametric and Nonparametric Statistical Procedures.
I am performing chi-square analyses for my dissertation data and am wondering about appropriate "post-hoc" analyses when chi-square values are significant. I read one resource (Sheskin, 2004) that says to look for standardized residuals whose absolute values are greater than 1.96 but another resource (Haberman, 1973) says adjusted residuals are the data to examine. I like the idea of using adjusted because several of them are greater than 1.96 and thus seem to identify sources of chi-square significance whereas all the standardized values are less than 1.96 and thus don't seem to identify sources of chi-square significance. Anyone know for sure which if these, if either, are the appropriate post-hoc data to examine? Thanks!
Brian.
Haberman, S.J. (1973). The analysis of residuals in cross-classified tables. Biometrics, 29, p. 205-220.
Sheskin, D.J. (2004). Handbook of Parametric and Nonparametric Statistical Procedures.