The problem with personality and intelligence testing

Editor’s note: If you would like to submit a letter for possible publication, please email it to editor@rationalist.com.au. See our Publishing Guidelines.
Dear Editor
After reading the leading article in the RSA Daily (15 August) on the study of the relationship between personality and intelligence, I’m unconvinced that psychological studies such as this qualify as science. The problem here lies in the methodology:
“It’s surprising that the methods used today are essentially the same as those used 100 years ago to study personality (i.e., mostly people responding to questions about how much a statement describes them) and cognitive ability (i.e., mostly people computing numbers, defining words, and manipulating shapes and patterns).” 
Where ‘personality’ or the ‘Big Five Personality Framework’ is concerned, I’m reminded of the old Myers Briggs Test, widely used in the eighties to assess the suitability of applicants for particular jobs, or the Keirsey Temperament Sorter. 
Using the same methodology mentioned above, these tests divided people into four personality types based on scored degrees of extraversion/introversion, sensation/intuition, thinking/feeling, perceiving/judging. Career paths, intellectual suitability and marriage partners were extrapolated from the individual’s score. 
The problem with these tests is that over a lifetime the scores do not remain the same. Life’s changes can change your score. For instance, previously, as an arts administrator with a socially...