Friday, September 16, 2016

Basic reasons to be critical of mainstream psychology: A short introduction

It's reasonable to assume that the authority and influence of modern psychology is as much due to the perception that it's objective and scientific as it is to the relevance of its subject matter. The problem is, when when it comes to the measurement of variables like intelligence, self esteem, happiness, motivation, etc., this objectivity is an illusion. This can be shown with the following argument:

1. The power and authority of psychologists rests on the perception that their conclusions are more objective or definitive than biased everyday value judgments. So, while someone may "feel that their friend is intelligent", a psychologist is seen as able to more definitively "measure their intelligence." (The same holds for similar cases involving beliefs, self-esteem, attitudes, personality, etc.)

2. Despite any suggestions to the contrary, psychological variables (unlike, e.g., length, weight, temperature, etc.,) are not literally measurable; they can only be measured figuratively.

3. The figurative measurement of psychological variables is only possible by relying on subjective, common-sensical assumptions about which form(s) of behavior exemplify a particular variable (e.g., in the case of intelligence, the behavioral acts that are assumed to indicate intelligence). Having done this, psychologists may then quantify the variable by doing one or both of the following ways...
    • Relying on subjective reflection to make a quantity out of something intrinsically non-quantitative, as in the case of a likert scale (e.g., on a scale of 1-5, how much is s/he able to think outside of the box?).
    •  Aggregating some set of target behaviors that exemplify a given variable, and then "measuring" the variable in terms of the number of these behaviors a person is judged to have performed.
IMPLICATIONS
The decisions described in (3) can only be made in terms of commonsense, everyday, subjective judgments. There is literally no other authority or basis on which to make these decisions.

Now, it could be argued that this criticism is too harsh. After all, the problems raised do not preclude the use of "measures" of things like intelligence, happiness, etc. in surveys or polls where their standardized form makes them useful for aggregating data and making comparisons across variables.  While there certainly are some valid and unproblematic uses of this type (Binet's use of the first intelligence test in the Paris school system is a good example), the previously identified problems are still very much relevant. The fundamental issue is that any findings reached with a measure that is constructed in the ways described above are only as definitive as the value judgments from which the measure was derived. No amount of care, standardization, validation or anything else can change this.

This limitation can be illustrated with the example of an intelligence test applied across different of groups of people. While the results of this might be described as group a performed significantly better than group b on this objective measure of intelligence, in reality the result would be somewhat more similar to a psychologist (acting indirectly via a written assessment) used their own judgment to evaluate the intelligence of a series of people and aggregated the results, revealing some group differences.