Nov 29 2009
When Psychology Goes Soft
Our society is currently undergoing this trend: Females are increasingly becoming the primary earners in households. This week I stumbled upon some research into the psychology of the phenomenon. While the write-up first showed promise, it ultimately disappointed me. The science was very soft.
The promise:
University of Missouri researcher Rebecca Meisenbach has found that women who take the role of lead breadwinner for their families experience both benefits and tensions. [source]
The disappointment:
In the qualitative study, Meisenbach interviewed 15 female breadwinners in professional occupations in the United States and had them describe their own experiences as breadwinners.
First there was the number. Just 15 subjects. But, more importantly, there was the type of research conducted. A qualitative study. This type of research is soft science if there ever were such a thing.
Here’s the beginning of the Wikipedia entry on qualitative research of the psychological variety:
In psychology, qualitative research has come to be defined as research whose findings are not arrived at by statistical or other quantitative procedures. Qualitative research is often said to be naturalistic. That is, its goal is to understand behaviour in a natural setting. Two other goals attributed to qualitative research are understanding a phenomenon from the perspective of the research participant and understanding the meanings people give to their experience. [bold mine]
To put it bluntly, it seems to me that this type of research is perfect for people averse to numbers and controls: the very things that give science it’s accuracy and precision. With qualitative research it is far too easy for the researchers’ own biases and preconceptions to plucked out of the vague cloud of data generated (unconsciously evoked?)
I don’t trust it. Qualitative research relies on the notion that people accurately perceive what is going on in their lives. Meanwhile, a mountain of research generated by strong scientific methods have highlighted how problematic this is. Furthermore, qualitative research is entirely non-replicable.
Sure, use it as a preliminary stage in data gathering. Employ qualitative methods to generate ideas about what might be the case. But once you have generated your set of “might be”s — hypotheses — to determine if they actually are I would insist on switching to more controlled, quantitative methods.
What did the above mentioned research study conclude?
Meisenbach found that the female breadwinners described six essential experiences: opportunities for control, independence, pressure and worry, valuing partner’s contributions, guilt and resentment, and ambition.
Has a bit of the feel of an astrological forecast, doesn’t it?
Are those “six essential experiences” truly what is going on? Maybe. But perhaps they are merely shared ways of thinking about what is going on (suggested/encouraged by the interviewer?). Without better controlled research, there is no way to know with any significant degree of confidence.
