Inconsistencies between mental fatigue measures under compensatory control theories, 2020, Muñoz-de-Escalona et al

Free full text: https://content.sciendo.com/view/jo...20-0006/article-10.2478-psicolj-2020-0006.xml

 

@Woolie

 

The sensation of fatigue has always been about the brain noting an increase in effort to maintain the same performance. The same is true of both muscular and cognitive exertion. The problem with this definition is it is hard to measure in a controlled manner. It is easy to measure a decline in cognitive performance (during maximal mental exertion) or maximal exertion, but this has little to do with the fatigue that people feel in their day to day lives (because people don't routinely exert maximally).

The fact that many neurologists and the like do not understand this, shows that they have been barking up the wrong tree.

 

It sounds very reasonable to me.

Yes, there is lots of evidence to suggest that people's perceptions of their level of cognitive fatigue and their performance don't match exactly, and there are many reasons for this mismatch that aren't "psychological" in the sense the BPSers use it. These authors are on exactly the right track. Humans are certainly very good at compensating for their limitations. At least in the short-term. Very demanding tasks require a high level of brain resources (good blood perfusion in the key regions), which requires high levels of autonomic arousal, and this is hard to sustain for long periods from a physiological point of view. Their pupil dilation thing in the paper is used as a proxy for those high levels of autonomic arousal, and is a neat measure.

One other reason for the observed mismatches between perceived and actual performance is poor sensitivity of the measures. There are two things here. The first is the sensitivity of a measure from a psychometric point of view - what are the measurement units, and what is the scale (ordinal, interval, etc), and how much of a change in effectiveness is needed to produce a single unit difference? Measures of accuracy in general seem to have a non-linear relationship with the underlying constructs they're measuring, and are often insensitive to small decrements in function.

Tasks that are designed to detect huge holes in people's heads will generally fail to detect the kinds of changes you see in even the most extreme fatigue. They're just not the right choices for the job.

The other thing is how "pure" the measure is. A measure that taps into multiple cognitive skills will be more resilient to decrements than one that mainly taps into a single skill. Also, measures that tap into a lot of crystallised knowledge (well-learned facts or skills) will tend to be less vulnerable to fatigue-related decrements that those that require a lot of on-the-spot maintenance and manipulation of information. Those things are more resource heavy.

There's a nice discussion of these issues here, at least as they relate to intelligence testing. Its a lovely piece, but a damned hard read, so only for the very dedicated!