[Bug]: SEM: Error is Node Stack Overflow

[kr-99]

JASP Version

0.19.0

Commit ID

No response

JASP Module

SEM

What analysis are you seeing the problem on?

while running the model for SEM analysis

What OS are you seeing the problem on?

Windows 10

Bug Description

jaspSem::SEM(
version = "0.19",
additionalFitMeasures = TRUE,
ave = TRUE,
errorCalculationMethod = "robust",
estimator = "dwls",
models = list(list(name = "Model 1", syntax = list(columns = list("JaspColumn_0_Encoded", "JaspColumn_1_Encoded", "JaspColumn_2_Encoded", "JaspColumn_3_Encoded", "JaspColumn_12_Encoded", "JaspColumn_13_Encoded", "JaspColumn_14_Encoded", "JaspColumn_7_Encoded", "JaspColumn_8_Encoded", "JaspColumn_9_Encoded", "JaspColumn_10_Encoded", "JaspColumn_16_Encoded", "JaspColumn_17_Encoded"), model = "certainty =~ JaspColumn_0_Encoded + JaspColumn_1_Encoded +JaspColumn_2_Encoded + JaspColumn_3_Encoded

graph =~ JaspColumn_12_Encoded + JaspColumn_13_Encoded + JaspColumn_14_Encoded

riskcalculation =~ JaspColumn_16_Encoded + JaspColumn_17_Encoded

numeracy =~ JaspColumn_7_Encoded + JaspColumn_8_Encoded + JaspColumn_9_Encoded + JaspColumn_10_Encoded

riskliteracy =~ certainty + graph + numeracy + riskcalculation

certainty ~ graph
graph ~ numeracy
numeracy ~ certainty
graph ~ riskcalculation
riskcalculation ~ numeracy
riskcalculation ~ certainty", modelOriginal = "certainty =~ certainty1 + certainty2 +certainty3 + certainty4

graph =~ graph1 + graph2 + graph3

riskcalculation =~ riskcalculation2 + riskcalculation3

numeracy =~ numeracy1 + numeracy2 + numeracy3 + numeracy4

riskliteracy =~ certainty + graph + numeracy + riskcalculation

certainty ~ graph
graph ~ numeracy
numeracy ~ certainty
graph ~ riskcalculation
riskcalculation ~ numeracy
riskcalculation ~ certainty"))),
naAction = "listwise",
pathPlot = TRUE,
pathPlotParameter = TRUE,
rSquared = TRUE,
standardizedEstimate = TRUE)
JASP SEM bug.docx

Expected Behaviour

I ran the same code in CFA and it modelled it correctly. But I am not able to run he analysis in SEM. I do not know what is the expected behavior.

Steps to Reproduce

1. I pasted the following in the Model, and it wont run.
   certainty =~ certainty1 + certainty2 +certainty3 + certainty4

graph =~ graph1 + graph2 + graph3

riskcalculation =~ riskcalculation2 + riskcalculation3

numeracy =~ numeracy1 + numeracy2 + numeracy3 + numeracy4

riskliteracy =~ certainty + graph + numeracy + riskcalculation

certainty ~ graph
graph ~ numeracy
numeracy ~ certainty
graph ~ riskcalculation
riskcalculation ~ numeracy
riskcalculation ~ certainty

Log (if any)

No response

More Debug Information

No response

Final Checklist

• I have included a screenshot showcasing the issue, if possible.
• I have included a JASP file (zipped) or data file that causes the crash/bug, if applicable.
• I have accurately described the bug, and steps to reproduce it.

[juliuspfadt]

Thanks for reporting the issue. I think without the data or the zipped .jasp file I cannot locate the problem.

[kr-99]

5country_riskliteracy_dataset_kr_sem2.zip

Here is the zipped .jasp file.

[juliuspfadt]

thanks. Could you try changing the scaling of the variables from nominal to ordinal?

[juliuspfadt]

I think that might be the error. At the moment all variable types are allowed but ideally the nominal variables would be transformed to ordinal (as is done in CFA). @boutinb do you know how to restrict the allowed types in the model window of SEM?

[kr-99]

Yes, in CFA it is being transformed into ordinal, by itself. Is it the issue of modelfit? I tried building the model from scratch, though.

[juliuspfadt]

The underlying R-package only takes ordinal or scale variables, not nominal. I dont know what you mean with model fit. I think this is a bug in the way the data is read in the SEM module.

[kr-99]

I tried changing to ordinal. Now the error is this,

his analysis terminated unexpectedly.

Error in if (nrow(originalSyntaxTable[originalSyntaxTable$op = ':=', : argument is of length zero

Stack trace
tryCatchOne(expr, names, parentenv, handlers[[1]])

doTryCatch(return(expr), name, parentenv, handler)

withCallingHandlers(expr = analysis(jaspResults = jaspResults, dataset = dataset, options = options), error = .addStackTrace)

analysis(jaspResults = jaspResults, dataset = dataset, options = options)

.semParameters(modelContainer, dataset, options, ready)

.semParameterTables(modelContainer[['results']][['object']][[1]], NULL, params, options, ready, modelContainer, dataset)

To receive assistance with this problem, please report the message above at: https://jasp-stats.org/bug-reports

[kr-99]

Hey guys, it runs now. Thanks. I changed the estimation options as well.

[juliuspfadt]

I think the issue now is, that the missing data handling is set to FIML, can you try setting it to "list wise deletion"

[juliuspfadt]

FIML and ordinal data dont work together. But the error message is really not helping

[kr-99]

Yes, I changed the estimation option to "DWLS" and the setting to "list wise deletion". Then it works.