Somebody on the forum highlighted the following to me. It has nothing specifically to do with ME/CFS. However, the odd person who is interested in psychology, or questionnaires and methods, might find it of interest .
user9876 has previously referred to some of these ideas about scales.
--------
A 2011 presentation by Matthew Hankins, Senior lecturer in Public Health at University of Southampton:
http://www.slideshare.net/mc_hankins/how-does-health-psychology-measure-up-9691694
--------
It questions how useful many questionnaires are that are used in health psychology. Also, questions how they are validated. It then discusses issues about whether certain questionnaires should be considered scales. It gets abstract so if you find this sort of stuff difficult, perhaps one to miss.
It mentions Rasch models, which I don't know much about, but that I was able to understand most of it without previous knowledge.
Possibly the most interesting slide to me was slide 17. It shows how a 3-factor solution doesn't necessarily mean three separate concepts are being measured; it may simply mean three different levels of severity are measured. It later goes on to show the particular scale mentioned, which has three factors, should be considered as unidimensional (or so the person argues and it seems convincing enough to me but I'm far from an expert).
user9876 has previously referred to some of these ideas about scales.
--------
A 2011 presentation by Matthew Hankins, Senior lecturer in Public Health at University of Southampton:
http://www.slideshare.net/mc_hankins/how-does-health-psychology-measure-up-9691694
--------
It questions how useful many questionnaires are that are used in health psychology. Also, questions how they are validated. It then discusses issues about whether certain questionnaires should be considered scales. It gets abstract so if you find this sort of stuff difficult, perhaps one to miss.
It mentions Rasch models, which I don't know much about, but that I was able to understand most of it without previous knowledge.
Possibly the most interesting slide to me was slide 17. It shows how a 3-factor solution doesn't necessarily mean three separate concepts are being measured; it may simply mean three different levels of severity are measured. It later goes on to show the particular scale mentioned, which has three factors, should be considered as unidimensional (or so the person argues and it seems convincing enough to me but I'm far from an expert).