February 10th, 2020 | Neuroscience vs Psychometrics. How To Research The Human Mind?

Neuroscience vs Psychometrics: Which Tool Is Better To Investigate the Human Mind?

For the whole last decade, I was fallen in love with Neuroscience. I first found out about the concept of investigating the human brain during my undergraduate studies, and from the very start, I knew that this was what I was going to do for a living. The human brain was such a challenging and fascinating concept!

I studied Physics, Mathematics, and Psychology in the College of Inter-Faculty Studies in Mathematics and Natural Sciences at the University of Warsaw at that time. In my undergraduate studies in Psychology, I chose a specialization in Psychometrics which was the most scientific among all the available specializations. 

Psychometrics is all about conducting behavioral research studies in humans: constructing tests and questionnaires, standardizing these tools on a representative (typically, adult) population, and validating them by comparing the results with results from other tools that might measure similar traits. 

The Common Critics of Psychometrics.

I remember myself running around the city with a bunch of papers and asking my friends from the Physics Faculty to fill in some tests for me. Most of them were not the happiest while filling in these tests and often complained that self-descriptive tools cannot give an objective measure for any personality traits or talents.

To some extent, I agree with them: it is very hard to make an objective test or questionnaire, for a few reasons. Firstly, it is hard to make a questionnaire in which the questions are not burdened with social approval to some extent. For instance, when asked “Do you learn faster than others?”, most people would answer “yes” as they have a biased perception of themselves. 

Secondly, not every question differentiates the population to enough extent—if 98% of the respondents say “yes” and only 2% say “no”, the question is not informative enough. Furthermore, the answers to multiple questions might strongly correlate which makes some of the questions redundant. 

Thirdly, most people tend to agree/conform with the statements they hear—if presented with 20 random statements and asked whether they identify with these traits/behaviors. the will respond “yes” to most of the points. 

Furthermore, there are massive gender differences when using the Likert scale (for instance, 1-7 where “1” means “fully agree” and “7” means “fully disagree”): men tend to give more extreme answers (more “1”s and “7”s) than women who usually prefer to tune down their answers (more “2”s and “6”s)

Lastly, test and questionnaire participants get fatigued while making decisions. When it comes to filling long questionnaires involving hundreds of points, they tend to lose concentration during the study, and at the end of the study, they are more careless about the answers than in the beginning.

To many neuroscientists, psychometrics is indeed a prehistoric research method to investigate the human brain which was the only option available before any neuroimaging methods were ever created. 

The whole concept behind neuroimaging is that, since you have a way of looking into the human brain at work, you no longer need to rely on self-descriptive, biased psychometric tools. 

This is also how I liked to think about brain research while tapping into this subject. It seemed to be the more modern, more objective, and more scientifically correct version of psychometric studies—and it all sounded just wonderful.

Afterthoughts After Spending Many Years in Neuroscientific research: Not As Cool As It Seems.

After these long years spent in neuroscience, I have a lot of doubts about how much you can learn about human minds from neuroimaging studies though. There are so many problems with today’s neuroscience! 

Firstly, what is the right level of depth to investigate the workings of the human mind? Large-scale neuroimaging methods such as fMRI or EEG cover very large neuronal populations interacting on the scale of the whole human brain while more acute methods such as optogenetics touch single neurons. It is quite likely that cognition works as a superposition of neuronal communication on many different spatial scales, and that none of the neuroimaging methods known today can give us a proper understanding of it. 

Secondly, neuroscience attracted a lot of public interest and public money. Due to official statistics, more than 100,000 researchers are working in neuroscience full-time around the globe. Neuroscience is a hybrid discipline of science that attracted many researchers who had previously worked in exact sciences, for example, in Mathematics, Physics or Computer Science, or in biomedical sciences, example, Biology, Biochemistry, Biotechnology, et cetera. 

With such an inflow of people, crazy competition and publication pressure have started. Together with this race for cool results, a lot of p-hacking has happened, and in a consequence, a lot of false-positive results appeared in the literature. In some branches of neuroscience, the amount of knowledge we have today is lower than ten years ago as it is no longer possible to tell what is true, and what is not. 

Conclusions: Neuroscience vs Psychometrics.

Overall, almost ten years of learning about the subject, and trying to contribute, I have a general feeling that I know less than in the beginning. And that if I want to know about how people feel or what they think, I better simply ask them.

And that’s my conclusion from my over ten-year-long odyssey through science; sometimes the underdog discipline actually gives better tools that lead to more useful outcomes than an overhyped, popular, “modern” discipline which is all around the media. 

Thus, I started talking to people more and looking into their brains less. In the future, probably most of my studies will be dedicated to building psychometric tools, and conducting psychometric studies rather than digging into the brains. I’ve been there, I’ve seen things, and I chose another way.