Table 3 COSMIN Summary of findings.

From: COSMIN review of the PANSS Marder factor solution and other factor models in people with schizophrenia

 

Summary or pooled result

Overall rating

Quality of evidence

Content validitya

   

M5M total score

Relevance rating: ±, Comprehensiveness rating: −, Comprehensibility rating: ± (overall ratings resulting from the PROM development study and the reviewer rating)

Inconsistent

Low (no content validity analyses available)

Structural validity

   

M5M total score

12/93 of the EFA derived models fulfill the criteria; CFI/NFI-range: 0.511–0.887 (total sample size: 12,025), 15/17 scalability assessments reject scalability

Insufficient

High (multiple analyses of very good quality; consistent results)

Positive factor

Unidimensional score

Indeterminate (3× (?))

High (multiple analyses of adequate quality; consistent results)

Negative factor

Unidimensional score; CFI-range: 0.973–0.982 (total sample size: 138)

Sufficient

High (multiple analyses of very good quality; consistent results)

Disorganized thought factor

Unidimensional score

Indeterminate (3× (?))

High (multiple analyses of adequate quality; consistent results)

Uncontrolled hostility/excitement factor

Unidimensional score

Indeterminate (3× (?))

High (multiple analyses of adequate quality; consistent results)

Anxiety/depression factor

Unidimensional score

Indeterminate (3× (?))

High (multiple analyses of adequate quality; consistent results)

Hypothesis testing

   

M5M total score

0 out of 5 hypotheses confirmed

Insufficient

Moderate (multiple analyses of doubtful quality, consistent results)

Positive factor

12 out of 17 hypotheses confirmed

Sufficient

Moderate (multiple very good analyses, inconsistent results)

Negative factor

13 out of 17 hypotheses confirmed

Sufficient

High (multiple very good analyses, consistent results)

Disorganized thought factor

9 out of 12 hypotheses confirmed

Sufficient

Moderate (multiple analyses of doubtful quality, consistent results)

Uncontrolled hostility/excitement factor

9 out of 12 hypotheses confirmed

Sufficient

Moderate (multiple analyses of doubtful quality, consistent results)

Anxiety/depression factor

10 out of 12 hypotheses confirmed

Sufficient

Moderate (multiple analyses of doubtful quality, consistent results)

Responsiveness

   

M5M total score

0 out of 5 hypotheses confirmed

Insufficient

High (multiple adequate analyses, consistent results)

Positive factor

2 out of 5 hypotheses confirmed

Inconsistent (2x (+), 3× (−); 60% (−))

None

Negative factor

3 out of 5 hypotheses confirmed

Inconsistent (3 (+), 2 (−); 60% (+))

None

Disorganized thought factor

3 out of 5 hypotheses confirmed

Inconsistent (3 (+), 2 (−); 60% (+))

None

Uncontrolled hostility/excitement factor

2 out of 5 hypotheses confirmed

Inconsistent (2x (+), 3× (−); 60% (−))

None

Anxiety/depression factor

2 out of 5 hypotheses confirmed

Inconsistent (2× (+), 3× (−); 60% (−))

None

  1. aThe domain content validity is independent of scale structure and therefore the content validity rating of the original three-subscale PANSS-30 was applicable to the M5M.