This is a unique time to be teaching and conducting psychological research in a university setting. In less than two decades, the development of new online services and diverse computational technologies (e.g., smart appliances, cars, homes and phones) has made it possible for researchers to study variables on an unprecedented scale. Since many people now produce continuous streams of data as they navigate the internet, interact with gadgets or communicate through social media, there is also the potential to study human behavior at vast levels of granularity with respect to time and geography. These developments have contributed to the rise of new interdisciplinary areas like “psychoinformatics” (Yarkoni, 2012), as well as growing awareness within traditional disciplines, like psychology, that “Big Data” technologies may have disruptive impacts. As others have speculated, the subfield of psychometrics may be in line to experience some of the most profound impacts from the new areas (Markowetz, Błaszkiewicz, Montag, Switala & Schlaepfer, 2014). Recently, however, our experiences teaching in the “Big Data” area has made us wonder if key impacts have not already taken considerable toll.
Although our home department is in psychology, for several years now we have been teaching in one of the new interdisciplinary graduate programs at our university connected to “Big Data.” The applied modelling and quantitative methods (AMOD) program was established in the early 1990s to be an interdisciplinary graduate program that drew faculty from a variety of disciplines who shared common interests in various statistical and mathematical modelling techniques (i.e., departments like biology, chemistry, computer science, economics, geography, mathematics, physics and psychology). More recently, the AMOD program created new course-based MSc. streams in “Big Data,” an initiative that has proven to be very popular with students. From a small “boutique” program enrolling a handful of new students each year, we have grown in just a few years to be the largest graduate program at our university. We are also the graduate program with the highest proportion of international students, attracting students with undergraduate backgrounds in computer science, software engineering, economics, finance and mathematics. Based on our routine tracking of graduates (and recent experiences teaching core courses with students), few of our students will go on to complete PhDs. The vast majority of our students are actively working towards starting “Big Data careers” in business and industry once they graduate.
One of the features of our graduate program is a capstone project the students complete in their last term of study. Working with our students as they carry out the projects has been highly revelatory, as we have been surprised at the number of projects that include various psychological variables. What has been even more surprising to observe is that interest in these variables does not come from a formal study of the relevant psychological literature. Only a tiny minority of our students have ever taken a psychology course, and virtually all of the prior work they cite in their projects comes from non-psychology journals.
To better understand this paradox, we undertook a brief citation analysis of relevant work. We searched for unique documents that used at least one key term of “machine learning,” “Big Data” or “algorithm,” with at least one of the following psychological key terms: “emotion,” “cognition” or “motivation” (The first three key words are quite essential for the Big Data area, while the latter three are pretty fundamental to the traditional discipline of psychology). Our database was documents in the ISI Web of Knowledge (WK; Thompson-Reuters, 2017) for the past decade. Fetscherin and Heinrich (2015) present evidence that this database is very useful when the goal is to conduct an interdisciplinary literature review. For the past decade (January 2009 to November, 2019), there were a total of 17,141 unique documents in the WK. Using the WK’s set of “research areas” for the source of the document (e.g., documents published in a “behavioral science” journal versus an “engineering journal,” etc.), the following list is the top 10 areas for the documents using our key terms (percentage of 17,141 documents listed in parentheses):
- Computer Science (70.1%)
- Mathematics (51.3%)
- Mathematical Computational Biology (35.0%)
- Engineering (33.3%)
- Science Technology (24.4%)
- Biochemistry (20.7%)
- Communication (18.0%)
- Neurosciences (17.9%)
- Biotechnology (17.7%)
- Behavioral Sciences (16.7%)
Although these are not independent categories, we were very surprised to see that “neurosciences” was only eighth on the list (17.9% of documents), with “behavioral sciences” 10th at (16.7%). “Psychology” sources were even lower at 12th place, accounting for only 15.1% of documents. These findings, as well as our experiences with our AMOD students, have led us to reflect on some of the challenges traditional psychology programs are going to face in the coming decade (especially those with a focus on psychometrics). Are we giving our psychology students sufficient training to thrive in the Big Data universe? As researchers, have we really embraced the new opportunities provided by Big Data technologies? We worry that psychology departments and programs may have inadvertently lost our hegemony in the study of core concepts in the discipline. While we rest on traditions and known principles, the world around us is changing. If we fail to change with it, do some of our subfields risk obsolescence?
About the authors
James D. A. Parker, PhD, is the director of the AMOD graduate program at Trent University. He has a long background in psychometrics as co-author of psychometric tools like the Coping Inventory for Stressful Situations (CISS), the BarOn Emotional Quotient Inventory (Youth Version (EQi:YV), and the 20-item Toronto Alexithymia Scale (TAS-20).
A. Geoffrey Crane completed an undergraduate in psychology and then an MSc. degree (thesis stream) from the AMOD program at Trent University. He currently teaches full time in the AMOD program.
References
Fetscherin, M., & Heinrich, D. (2015). Consumer brand relationships research: A bibliometric citation meta-analysis. Journal of Business Research, 68(2), 380-390. doi.org/10.1016/j.jbusres.2014.06.010
Markowetz, A., Błaszkiewicz, K., Montag, C., Switala, C., & Schlaepfer, T. E. (2014). Psycho-informatics: Big data shaping modern psychometrics. Medical hypotheses, 82(4), 405-411. doi.org/10.1016/j.mehy.2013.11.030
Thompson-Reuters (2017). Web of Science. https://www.webofknowledge.com/opens in new window
Yarkoni, T. (2012). Psychoinformatics: New horizons at the interface of the psychological and computing sciences. Current Directions in Psychological Science, 21(6), 391-397. doi.org/10.1177/0963721412457362
