estat gof, stats(all) not reporting values after SEM

[Kiley Davan]

I am analyzing data using an SEM and for certain models the RMSEA and CFI/TFI are not being calculated after using the command estat gof, stats(all). In particular, I had a final model (model1) for which estat gof, stats(all) calculated and reported all fit statistics. However, when I remove some paths (model2), estat gof, stats(all) no longer reports all fit statistics. For both models, all variables starting with dum_ and the variable 'harvested' are binary while the rest are ordinal. Any input would be much appreciated. Please see below for my code.

Model1

sem (Expectation -> gameseen_satisfaction, ) (Expectation -> overallsat, ) (dum_info_own -> Expectation, ) (dum_info_TWRA -> Expectation, ) (dum_info_club -> Expectation, ) (harvest_sat -> overallsat, ) (dum_info_online -> Expectation, ) (dum_info_mag -> Expectation, ) (dum_info_friends -> Expectation, ) (dum_info_news -> Expectation, ) (Constraints -> bg_con, ) (Constraints -> interest_con, ) (Constraints -> success_con, ) (Constraints -> cost_con, ) (Constraints -> game_pops_con, ) (Constraints -> quality_con, ) (Constraints -> crowded_con, ) (Constraints -> landowners_con, ) (Constraints -> regulations_con, ) (Constraints -> elders_con, ) (Constraints -> recruitment_con, ) (Constraints -> people_con, ) (Constraints -> commitments_con, ) (Constraints -> overallsat, ) (success_con -> Expectation, ) (success_con -> gameseen_satisfaction, ) (game_pops_con -> Expectation, ) (game_pops_con -> gameseen_satisfaction, ) (harvested -> harvest_sat, ) (Motivations -> food_motivation, ) (Motivations -> seeing_motivation, ) (Motivations -> skills_motivation, ) (Motivations -> challenge_motivation, ) (Motivations -> trophy_motivation, ) (Motivations -> solitude_motivation, ) (Motivations -> fincancial_motivation, ) (Motivations -> friends_motivation, ) (Motivations -> control_motivation, ) (Motivations -> service_motivation, ) (Motivations -> tradition_motivation, ) (Motivations -> teaching_motivation, ) (Motivations -> overallsat, ) (dum_info_social -> Expectation, ) (gameseen_satisfaction -> harvest_sat, ) (gameseen_satisfaction -> overallsat, ), covstruct(_lexogenous, diagonal) cov(_lexogenous*_oexogenous@0) latent(Constraints Motivations ) cov( e.bg_con*e.landowners_con e.success_con*e.game_pops_con e.cost_con*e.regulations_con e.game_pops_con*e.quality_con e.game_pops_con*e.crowded_con e.quality_con*e.success_con e.crowded_con*e.quality_con e.elders_con*e.recruitment_con e.people_con*e.commitments_con Motivations*Constraints e.seeing_motivation*e.solitude_motivation e.skills_motivation*e.challenge_motivation e.challenge_motivation*e.solitude_motivation e.challenge_motivation*e.fincancial_motivation e.friends_motivation*e.tradition_motivation e.friends_motivation*e.teaching_motivation e.control_motivation*e.service_motivation e.teaching_motivation*e.tradition_motivation e.gameseen_satisfaction*e.success_con) nocapslatent

estat gof, stats(all)

Model2

sem (Expectation -> gameseen_satisfaction, ) (Expectation -> overallsat, ) (dum_info_own -> Expectation, ) (dum_info_TWRA -> Expectation, ) (dum_info_club -> Expectation, ) (harvest_sat -> overallsat, ) (dum_info_online -> Expectation, ) (dum_info_mag -> Expectation, ) (dum_info_friends -> Expectation, ) (dum_info_news -> Expectation, ) (Constraints -> bg_con, ) (Constraints -> interest_con, ) (Constraints -> success_con, ) (Constraints -> cost_con, ) (Constraints -> game_pops_con, ) (Constraints -> quality_con, ) (Constraints -> crowded_con, ) (Constraints -> landowners_con, ) (Constraints -> regulations_con, ) (Constraints -> elders_con, ) (Constraints -> recruitment_con, ) (Constraints -> people_con, ) (Constraints -> commitments_con, ) (Constraints -> overallsat, ) (harvested -> harvest_sat, ) (Motivations -> food_motivation, ) (Motivations -> seeing_motivation, ) (Motivations -> skills_motivation, ) (Motivations -> challenge_motivation, ) (Motivations -> trophy_motivation, ) (Motivations -> solitude_motivation, ) (Motivations -> fincancial_motivation, ) (Motivations -> friends_motivation, ) (Motivations -> control_motivation, ) (Motivations -> service_motivation, ) (Motivations -> tradition_motivation, ) (Motivations -> teaching_motivation, ) (Motivations -> overallsat, ) (dum_info_social -> Expectation, ) (gameseen_satisfaction -> harvest_sat, ) (gameseen_satisfaction -> overallsat, ), covstruct(_lexogenous, diagonal) cov(_lexogenous*_oexogenous@0) nolog latent(Constraints Motivations ) cov( e.bg_con*e.landowners_con e.success_con*e.game_pops_con e.cost_con*e.regulations_con e.game_pops_con*e.quality_con e.game_pops_con*e.crowded_con e.quality_con*e.success_con e.crowded_con*e.quality_con e.elders_con*e.recruitment_con e.people_con*e.commitments_con Motivations*Constraints e.seeing_motivation*e.solitude_motivation e.skills_motivation*e.challenge_motivation e.challenge_motivation*e.solitude_motivation e.challenge_motivation*e.fincancial_motivation e.friends_motivation*e.tradition_motivation e.friends_motivation*e.teaching_motivation e.control_motivation*e.service_motivation e.teaching_motivation*e.tradition_motivation) nocapslatent

estat gof, stats(all)

[Kiley Davan]

I was able to figure out what was causing this issue and wanted to post it just in case someone else has a similar problem. Removing the covariance path between the success_con and game_pops_con variables resolved my issue with the fit statistics not reporting. However, I am not sure exactly why this covariance path was causing a problem in the first place (especially since both the success_con and game_pops_con variables have covariance paths to other variables). So, if you are having a similar issue to this, I suggest you rebuild your model (starting with a model you know works) and add in other variables/paths one at a time to identify what is causing the issue.