Hello community,
My dataset is only composed of binary variables (0/1) with 49 observations. When I apply the PLS-SEM method using the SEMinR package, I encounter an issue: I can't run the bootstrap procedure on my model due to the low variances in the data. However, I really need a way to test the significance of the model estimates.
Here is my code:
<measurement_model <- constructs(
composite("Barriers", multi_items("", c("Data.ownership.conflicts", "Lack.of.accessibility", "Skills.gap", "Legislative.constraints")), weights = mode_B),
composite("Opportunities", multi_items("", c("Accelerating.innovation", "Reducing.costs", "Climate.change.adaptation.and.mitigation", "Job.creation", "Improving.effectiveness")), weights = mode_B),
composite("Solutions", multi_items("", c("Standardisation", "Empowering.individuals", "Contractual.models", "Legal.intervention", "Public.policy", "Emerging.technologies", "Facilitating.data.sharing.and.access")), weights = mode_B),
composite("Challenges", multi_items("", c("Risks.of.discrimination", "Ethical.considerations", "Privacy.concerns", "Security.threats", "Environmental.costs", "Public.acceptance.and.governance.issues")), weights = mode_B),
composite("Drivers", multi_items("", c("Interoperability", "Legislation", "Technological.enablers")), weights = mode_B)
)
structural_model <- relationships(
paths(from = "Drivers", to = c( "Opportunities")),
paths(from = "Solutions", to = c("Challenges","Barriers"))
)
pls_model <- estimate_pls(data = merged3_df,
measurement_model = measurement_model,
structural_model = structural_model,
)>
Thank you!