Improving the Measurement of Functional Somatic Symptoms With Item Response Theory, 2020, Acevedo-Mesa et al

[Andy]

More than 40 questionnaires have been developed to assess functional somatic symptoms (FSS), but there are several methodological issues regarding the measurement of FSS.

We aimed to identify which items of the somatization subscale of the Symptom Checklist–90 (SCL-90) are more informative and discriminative between persons at different levels of severity of FSS. To this end, item response theory was applied to the somatization scale of the SCL-90, collected from a sample of 82,740 adult participants without somatic conditions in the Lifelines Cohort Study. Sensitivity analyses were performed with all the participants who completed the somatization scale.

Both analyses showed that Items 11 “feeling weak physically” and 12 “heavy feelings in arms or legs” were the most discriminative and informative to measure severity levels of FSS, regardless of somatic conditions. Clinicians and researchers may pay extra attention to these symptoms to augment the assessment of FSS.

Open access, https://journals.sagepub.com/doi/10.1177/1073191120947153

 

[rvallee]

The item response theory (IRT), also known as the latent response theory refers to a family of mathematical models that attempt to explain the relationship between latent traits (unobservable characteristic or attribute) and their manifestations (i.e. observed outcomes, responses or performance).

So, mathemagics? If you torture the data long enough, it will eventually give you an answer. It won't be a relevant answer but you will be able to report that you have an answer.

This is due to the large proportion of the participants reporting not at all (0) in all items, reducing the probability of choosing higher answer options and inflating the threshold parameters

Sounds like either their sample is bad, their questions or bad, or both. Anyway this is a super category with many different things, you'd get a bunch of weird answers as well if you put it any number of unrelated diseases and waterboarded the data a little. Junk drawer indeed.

 

[Snowdrop]

It sounds like they might know something about performing data analysis. But it seems to me that this is inadequate for bringing to light any understanding of these health issues which would require a more nuanced understanding of what might be valid to start with IMO.