Psychological Treatment Symptom Data for Predicting Outcomes with Latent Trait Scores

6,464 patient records from a 12-week psychological treatment program, used to compare predictive models using traditional sum scores versus Rasch-derived latent trait scores. The dataset contains self-rated symptom data from three questionnaires collected over the first 4 weeks to predict post-treatment outcomes. Nils Hentati Isacsson published this data on figshare in March 2026 under a CC-BY-4.0 license.

Use Cases

• Comparing predictive performance of latent trait scores versus ordinal sum scores based on the described psychometric analysis.
• Training regression or random forest models to predict post-treatment symptom outcomes based on early-treatment data.
• Evaluating the impact of questionnaire length reduction (to 23% of the original scale) on prediction accuracy.
• Studying measurement invariance and unidimensionality in psychological assessment using Rasch Measurement Theory principles.

Strengths

• Data from 6,464 patients undergoing a standardized 12-week treatment program.
• Includes results from three different self-rated symptom questionnaires.
• Prediction models were validated using nested cross-validation and multiple imputation for missing data.
• Published under a permissive CC-BY-4.0 license for reuse.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the specific data files is unknown, which may limit suitability assessment.
• The 606.5 KB file size suggests a limited scope, likely containing aggregated results or a subset of the full analysis data.

Provenance

Source

figshare

Collection Method

Self-rated symptom data collected from patients during psychological treatment.

Time Range

Data collected over the first 4 weeks of a 12-week treatment program.

Freshness

Last updated 2026-03-18 07:45:59; freshness should be verified.

Geography

null