I’m comparing two two-level multilevel models – fit0 and fit1 – and want to know if the difference slope between the two models is significant. As the approach by Clogg et al. (1995) does not allow nested data, I use GEE models and the compCoef function of the package “geepack” (https://cran.r-project.org/web/packages/geepack/geepack.pdf). I did run my models and employed the compCoef function as followed:

fit0 <- geese(mom_MH_CS ~ 1 + CR_unpl_and_pl_MC + nr_entry + day_sitnumberneg_MC + day_sitnumberpos_MC + grandmean_NA_unpl + grandmean_PA_pl + trait_gender, id = vp_initials, data = mds_fakenames, corstr = "un")

fit1 <- geese(mom_MH_CS ~ 1 + CR_unpl_and_pl_MC + nr_entry + day_sitnumberneg_MC + day_sitnumberpos_MC + grandmean_NA_unpl + grandmean_PA_pl + trait_gender, id = vp_initials, data = mds, corstr = "un")

compCoef(fit0, fit1)

However, I have problems interpreting the variance table provided by R. namely, I do not know for certain how to obtain the standard deviation or standard error required to obtain the z-score. I didn’t find an answer for it neither in the pdf-guideline nor in the related paper of Yang et al. (2001; https://www.jstor.org/stable/41999419). Thus, I would like to ask you which of the following values you would choose for the error variance of CR_unpl_and_pl_MC – or if you would say it isn’t stated at all.

Thank you very much for your assistance

~Erefin

And now here’s the aforementioned table:

$variance
                         beta_(Intercept) beta_CR_unpl_and_pl_MC   beta_nr_entry beta_day_sitnumberneg_MC beta_day_sitnumberpos_MC
beta_(Intercept)           0.000081890472        -0.000001410043  0.000005539705           0.000048402922           0.000048411871
beta_CR_unpl_and_pl_MC    -0.000001410043         0.000227988689 -0.000004134354          -0.000015684815          -0.000056329257
beta_nr_entry              0.000005539705        -0.000004134354  0.000001137281           0.000006325759           0.000006822137
beta_day_sitnumberneg_MC   0.000048402922        -0.000015684815  0.000006325759           0.000044558401           0.000045579655
beta_day_sitnumberpos_MC   0.000048411871        -0.000056329257  0.000006822137           0.000045579655           0.000057384384
beta_grandmean_NA_unpl     0.000029022541        -0.000009342169  0.000004161044           0.000026915631           0.000028121915
beta_grandmean_PA_pl      -0.000080381837         0.000027915699 -0.000010935171          -0.000071103276          -0.000074670748
beta_trait_gender         -0.000143420550         0.000047145153 -0.000017997632          -0.000117529393          -0.000123421272
                         beta_grandmean_NA_unpl beta_grandmean_PA_pl beta_trait_gender
beta_(Intercept)                 0.000029022541       -0.00008038184    -0.00014342055
beta_CR_unpl_and_pl_MC          -0.000009342169        0.00002791570     0.00004714515
beta_nr_entry                    0.000004161044       -0.00001093517    -0.00001799763
beta_day_sitnumberneg_MC         0.000026915631       -0.00007110328    -0.00011752939
beta_day_sitnumberpos_MC         0.000028121915       -0.00007467075    -0.00012342127
beta_grandmean_NA_unpl           0.000021692294       -0.00004836462    -0.00007327497
beta_grandmean_PA_pl            -0.000048364617        0.00012741010     0.00020168516
beta_trait_gender               -0.000073274970        0.00020168516     0.00035669130