GSEM (path analysis): initial values not feasible

[Radhouene DOGGUI]

Hi All,

Please I am trying to run this model but I get an error message that I am not able to resolve. I would appreciate any help.

Thanks


gsem (i.sex -> _zbfa, family(gaussian) link(identity)) (i.m_edu_3cat -> SES, ) (i.ecproxy3_w -> SES, ) (i.f_edu_3cat -> SES, ) (i.actm_c3 -> SES, ) (i.actp_c3 -> SES, ) ///
(SES-> _zbfa, family(gaussian) link(identity)) (age -> _zbfa, family(gaussian) link(identity) ) ///
(SES -> i.affss_cat1, ) (SES -> i.mvpa_cat, ) (SES -> i.pg_knowledge_3cat, ) ///
(Child_food_intak -> _zbfa, family(gaussian) link(identity)) ///
(sum_dd_meats_fish-> Child_food_intak, ) (sum_dd_chick_meat -> Child_food_intak, ) (sum_dd_eggs -> Child_food_intak, ) (sum_dd_milk -> Child_food_intak, ) (sum_dd_vita_veg -> Child_food_intak, ) (sum_dd_fat -> Child_food_intak, ) (sum_dd_salted_snack -> Child_food_intak, ) (sum_dd_cereals_tubers -> Child_food_intak, ) (sum_dd_sweets -> Child_food_intak, ) (sum_dd_sweetened_bev -> Child_food_intak, ) (sum_dd_fruits_veg_mod -> Child_food_intak, ) ///
(i.affss_cat1->_zbfa, family(gaussian) link(identity)) (i.affss_cat1 -> Child_food_intak, ) ///
(Parent_fedding_behav -> _zbfa, family(gaussian) link(identity)) (Parent_fedding_behav -> Child_food_intak, ) (Parent_fedding_behav -> rest_score, ) (Parent_fedding_behav -> press_score, ) (Parent_fedding_behav -> monit_score, ) (Parent_fedding_behav -> model_score, ) (Parent_fedding_behav -> conc_score, ) ///
(i.mvpa_cat -> _zbfa, family(gaussian) link(identity)) ///
(i.pg_knowledge_3cat -> Child_food_intak, ) (i.pg_knowledge_3cat -> Parent_fedding_behav, ) (i.pg_knowledge_3cat -> mother_bmi, ) (i.pg_knowledge_3cat -> father_bmi, ) ///
(Parent_nut_status -> _zbfa, family(gaussian) link(identity)) ///
(mother_bmi -> Parent_nut_status, ) (father_bmi -> Parent_nut_status, ) ///
(M1[Sclid] -> _zbfa, family(gaussian) link(identity)) ///
(food_sch_env -> _zbfa, family(gaussian) link(identity)) (food_sch_env -> Child_food_intak, ) ///
(i.scl_outlet_binary -> food_sch_env, ) (i.scl_outlet_num -> food_sch_env, ) (i.fp5 -> food_sch_env, ) (i.fp9 -> food_sch_env, ) ///
(i.governorates_id -> _zbfa, family(gaussian) link(identity)) ///
(i.food_reward -> food_sch_env, ) (tea_avgsknow -> food_sch_env, ) (i.ter_de_unhealthy -> food_sch_env, ) (i.ter_de_healthy -> food_sch_env, ) (i.ter_de_ads_processedultra_foods -> food_sch_env, ) (i.ter_de_ads_unprocessed_foods -> food_sch_env, ), ///
covstruct(_lexogenous, diagonal) latent(SES Child_food_intak Parent_fedding_behav Parent_nut_status M1 food_sch_env ) ///
cov( e.SES*e.Child_food_intak e.SES*e.Parent_fedding_behav e.rest_score*e.press_score e.rest_score*e.monit_score e.rest_score*e.model_score e.press_score*e.monit_score ///
e.press_score*e.model_score e.monit_score*e.model_score e.conc_score*e.rest_score e.conc_score*e.press_score e.conc_score*e.monit_score e.conc_score*e.model_score) nocapslatent


note: 4.scl_outlet_num omitted because of collinearity

Fitting fixed-effects model:

Iteration 0: log likelihood = -38614.653 (not concave)
Iteration 1: log likelihood = -38241.079
Iteration 2: log likelihood = -38221.876
Iteration 3: log likelihood = -38015.545
Iteration 4: log likelihood = -38009.616
Iteration 5: log likelihood = -38009.561
Iteration 6: log likelihood = -38009.561

Refining starting values:

Grid node 0: log likelihood = .
Grid node 1: log likelihood = .
Grid node 2: log likelihood = .
Grid node 3: log likelihood = .
Grid node 4: log likelihood = .
Grid node 5: log likelihood = .
Grid node 6: log likelihood = .
Grid node 7: log likelihood = .
Grid node 8: log likelihood = .
Grid node 9: log likelihood = .
.
.
.

Grid node 724: log likelihood = .
Grid node 725: log likelihood = .
Grid node 726: log likelihood = .
Grid node 727: log likelihood = .
Grid node 728: log likelihood = .
Grid node 729: log likelihood = .
(note: Grid search failed to find values that will yield a log likelihood value.)

Fitting full model:

initial values not feasible
r(1400);

end of do-file

r(1400);