"The first official detailed investigation of the validity of psychological measurement from beyond its professional ranks was conducted – under the auspices of the British Association for the Advancement of Science – by the Ferguson Committee in 1932. The non-psychologists on the committee concluded that there was no evidence to suggest that psychological methods measured anything, as the additivity of psychological attributes had not been demonstrated..." by Dr Hugh Morrison
https://paceni.wordpress.com/tag/holders-seven-axioms/
Above conclusion touches the essence of psychological measurement. How do we understand Total Score, General Total Score (Ability Measure), Item Response Theory and Conjoin measurement with this conclusion?
You might want to review the literature from the last 85 years. Perhaps you could begin with the concepts developed by Percy Bridgman and S.S. Stevens.
Hi, Stephen, The reason for me to pop up this question is to clarify the differences among Total Score, General Total Score, IRT and Conjoin Measurement because, since this conclusion was made by a group of Scientists 85 years ago in UK, I did not know any serious discussion on this conclusion. Again, this conclusion is consistent with Probability, Entropy, Fisher Information and correlation (R squared) in linear space etc. and therefore, I believe that the psychological measurement must also comply with this conclusion because they are all called the measure ( no matter what they measure). Nan Kong
Hi, Jan Rasmus Böhnke, Thank you for those articles.
The general version of "Additivity" is called "Decomposition Theory" because "Decomposition Theory" must be able to decompose the total object into those "overlapped" or "interactive" small parts while "Additivity" usually means to decompose total object into those small parts without "overlap" or "interaction" (therefore, "Additivity" is usually true under the independent condition). I would prefer the following conclusion:
"No evidence to suggest that psychological methods measured anything, as the decomposition theory of psychological attributes had not been demonstrated ..."
Some articles suggest Item Response Theory (IRT) or Additive Conjoint Measurement (ACM) while some won't, but no one seriously introduce the concept of "Decomposition theory". All those well-defined measure must be decomposable such as Probability, Entropy, Fisher Information or those measures for energy, mass or force In physics (no "overlap" for most attributes in Physics -- ACM works for these). But looking back into those psychological measurements, no "decomposition theory" is demonstrated or proved. The facts are that (1) ACM had no its decomposition theory because ACM is based on "independence" (without "independence", the cancellation rules are not true); (2) IRT has no additivity in its theoretical structure. Therefore, no evidence to suggest that IRT measures anything while ACM is not able to measure those psychological attributes with interactions.
Total Score has its additive structure and General Total Score has its "Decomposition Theory". Therefore, it is very possible for General Total Score to be the solution to the conclusion made by a group of scientists more than 80 years ago in UK.
- Badges
- Science topic
- Similar topics
- Psychology
- Psychometrics