Back

Development and psychometric evaluation of The Index of Myalgic Encephalomyelitis Symptoms (TIMES) Part I: Rasch Analysis and Content Validity

Horton, M. C.; Tyson, S. F.; Fleming, R.; Gladwell, P.

2026-02-17 health systems and quality improvement
10.64898/2026.02.16.26346394 medRxiv
Show abstract

ObjectiveTo develop and psychometrically evaluate an assessment of symptoms in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) MethodsAn initial symptom list was devised from the relevant literature with the patient and clinician advisory groups. An online survey with 85 symptom items in eight domains was completed by people with ME/CFS. Each item had two response structures (assessing symptom frequency and severity on five-point scales). Rasch analysis assessed each domain for unidimensionality, targeting, internal reliability, item fit and local dependency. ResultsSurvey data (n=721) indicated various item anomalies and inter-item dependencies, leading to item re-formatting or removal. The frequency and severity-based responses broadly replicated each other, and a four-point response format appeared more appropriate than a five-point response format. Following Rasch-based scale amendments, a revised version with a single four-point response format was re-administered to test the modifications. Validation data (n=354) showed the modified scale had an improved response structure and functionality across all domains, satisfying Rasch model assumptions. Additionally, domain-level super-items allowed for a summated total score along with sub-scales summarising neurological and autonomic symptoms, again satisfying Rasch model assumptions. ConclusionsThe Index of ME Symptoms (TIMES) and its associated sub-scales and domain scales are stable, valid assessments of symptoms in ME/CFS.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.