lavaan ERROR: unknown ov.types:haven_labelled

[kiki ekiawan]

Hi everyone, 

I have a problem encountered while running SEM in R. I saw a previous thread about it, but the problem cause a little bit different. in my case, I was able to run my model successfully when I treated my variables with numerical data, however, some of my variables were categorical (ordinal and nominal) and it's resulted error as follows: 

Error in lav_samplestats_step1(Y = Data, wt = wt, ov.names = ov.names,  :
  lavaan ERROR: unknown ov.types:haven_labelled
In addition: Warning message:
In lav_data_full(data = data, group = group, cluster = cluster,  :
  lavaan WARNING: exogenous variable(s) declared as ordered in data: Age_group Female_male Income Education

I did try to erase my value label from spss dataset and imported back, and I also use: sapply(diss_data, haven::zap_labels)  and sapply(diss_data, haven::as_factor). But still they are not successful to run.

This is my syntax: 

testmodel2DatasetOCT_SES<- '
# measurement model
RegulatingESy1=~ be_air_quality_lik + be_water_quality_erosion_lik+
  be_temperature_reduction_lik+ be_space_for_plants_animals_lik+
  be_noise_reduction_lik+ be_carbon_storage_climate_lik+
  be_natural_hazards_lik
ProvisioningESy2=~ be_other_products_lik+be_wood_lik+be_fuel_lik+be_hunting_lik
CulturalESy3 =~ be_human_health_wellbeing_lik+be_education_lik+be_cultural_emotional_spiritual_lik+
  be_jobs_economy_lik+ be_recreation_sport_lik+be_beauty_lik

#regression
RegulatingESy1~ Age_group+Female_male+Income+Education+Rurality+UN_geoscheme
ProvisioningESy2 ~ Age_group+Female_male+Income+Education +Rurality+UN_geoscheme
CulturalESy3 ~ Age_group+Female_male+Income+Education +Rurality+UN_geoscheme '

fitModeltest2SES_FES<- sem(testmodel2DatasetOCT_SES, data = DatasetCH_K2_exported_OCT23)

I hope to gain insight from the advanced people here. 

Thanks so much, 

Eki

[Yago Luksevicius de Moraes]

Hi, Eki,

lavaan cannot use the thresholds of uncorrelated ordinal variables.
The easiest way to solve this problem is to make these variables correlate with something. For instance, add the line

Age_group~~Female_male+Income+Education

in your model may solve the problem.

Best regards,
Yago

​