How WEIRD is music psychology?

A recent criticism of psychological research is that it is WEIRD. Yes, it is a bit weird that psychologists spend most of our days subjecting participants to hours-long tasks where they should press the key on the left if they hear ‘BLEEP’ and the key on the right if they hear ‘BLOOP’, but the ‘WEIRD’ I’m referring to is actually an acronym that stands for the fact that most psychological research is conducted on participants who are from Western, Educated, Industrialized, Rich, and Democratic societies. More specifically, a disproportionate amount of this research has been conducted on university students, primarily in the US, in exchange for course credit or occasionally a small fee (usually only enough to afford a small Frappucino at Starbucks). It has been suggested that  the WEIRD phenomenon may be “skewing our understanding of human behaviour”, as researchers have shown that WEIRD participants differ from those from non-WEIRD backgrounds in their performance on a variety of commonly used tasks, such as those measuring moral decision making and fairness, reasoning style, and perception of visual illusions.

Although the WEIRD problem has been discussed widely in relation to psychology research in general, its effects within the domain of music psychology have not been examined in detail. In order to begin to understand the extent to which our field might also be affected by WEIRD sampling biases, I collected information about the participant samples of 97 studies that examined the impact of background music on concurrent task performance (including tasks such as reading, driving, drinking, and shopping). This is a popular area within music psychological research, due to the implications listening to background music might have on productivity and information retention during studying, consumer behaviour in shops and restaurants, achievement during sporting events, and performance on many other everyday tasks. I made use of a list of studies on background music listening compiled for a meta-analysis by Kämpfe, Sedlmeier, and Renkewitz (2011). I searched for the 97 studies included in this meta-analysis and made note of the country in which each study was conducted, the study’s sample size and age, and any other relevant information about the background of the participant sample. (It is interesting to note that the WEIRD acronym doesn’t explicitly refer to the fact that much previous psychology research also displays an age bias due to primarily recruiting undergraduate students as participants. However, participant age will also be considered here as a related concept).

Of the 97 studies, I was unable to access the full paper for 20 studies in an internet search. (This highlights another current hot topic in academic research- the accessibility of papers and the move towards a more open access and transparent publishing system. Stay tuned for more on this in a future post!) However, I was able to extract most of the necessary information from the abstract for 13 of these 20 studies, so was able to proceed with information from 90 studies.

The first point of interest was the country in which each study was conducted. Figure 1 below shows the countries in which the studies took place. It is clear from this figure that the WEIRD phenomenon of overrepresentation of Western, industrialised, rich, and democratic countries is also present within music psychological research, with the most studies in this sample coming from the US (43 studies) and the UK (21 studies). Four studies each were conducted in Canada, Germany, and Japan; only one or two studies were conducted in all other countries highlighted on the map. Although there is a chance that some participants in each sample came from a country other than that in which the study was conducted, it can be assumed that the majority of the sample was not foreigners, and in some cases it was explicitly stated in a paper that foreign participants were excluded from the study due to language barriers or potential cultural differences.

The next data I looked at were qualitative descriptions of the sample used for each study, if any such information was given. 33 studies (37%) referred to the use of a sample of “undergraduate students” and 26 studies (29%) referred to the use of a sample of “university students”, “college students”, or “undergraduate and graduate students” (see Figure 2). This adds the ‘E’ to ‘WEIRD’, demonstrating that most of these studies focused primarily on people who have pursued or are currently pursuing a higher education degree. This also suggests the presence of an age bias towards participants in their late teens and early- to mid-twenties. An interesting point to note here, however, is that 12 of the studies (13%) were field studies conducted in shops, restaurants, malls, etc., in which the effects of background music on behaviour were studied in everyday contexts, primarily in terms of consumer behaviour (e.g. does a certain type of music make you spend more time or money in a shop, or eat or drink more at a restaurant?). This represents an effort to extend at least some aspects of this vast research area into more ecological contexts, which can also extend the target demographic to a wider range of ages and socioeconomic backgrounds than typical undergraduate students.

Finally, I collected information on the age range and/or mean age of the participant samples. One surprising finding from this step was that 29 of the studies did not provide any description of the age range or mean age of the sample. This might be due to different reporting conventions in different fields (e.g. several of these papers were published in marketing or sports science journals) or the changing of conventions over time (the studies span a wide range of publication dates, with the earliest being published in 1931). However, the vast majority of the studies that did not report age data reported their sample to be composed of undergraduates or university students, which suggests that most of their sample fell within the range of around 18 to 21 years of age. For studies that did report mean age data, the mean age of the participant sample across all studies, weighted by sample size, was 22.6 years. Notable  exceptions of studies that did not primarily focus on university students were: 1) several of the field studies that sampled participants from public places, such as supermarkets and malls, and thus included many participants in their 30s, 40s, 50s, and 60s, 2) four studies that focused on sporting performance and therefore recruited participants who regularly engaged in sports, such as karate class attendees and a netball team, and 3) two studies that focused specifically on older adult volunteers from the 60+ age category.

The results of this short exploration suggest that music psychology is in many ways just as susceptible to WEIRD biases as other psychological research. One positive point to note is that at least some of the research questions of interest within this particular field are conducive to study in more naturalistic contexts, such as shops or restaurants, in which one is likely to find participants who are from a wider range of backgrounds than typical university students. However, there are many other questions that music psychologists ask that are less easily adapted to field research and much better suited to study within a controlled lab setting. Considering that most of these “controlled lab settings” have been set up in WEIRD countries, it will not be a quick or easy fix to begin to conduct more research with non-WEIRD participants. However, it does appear that more efforts are beginning to be made to address this problem within our field. Within the Music & Science Lab at Durham University, we are presently running a project on Interpersonal Entrainment in Music Performance. This project combines the knowledge of music psychologists with ethnomusicologists and computer engineers in order to provide cross-cultural insights into the phenomenon of musical entrainment. This is being achieved in part by making use of videos recorded by ethnomusicologists of musical performances from across the world, in such locations as India, Cuba, and Africa, and bringing these into our “controlled lab setting” in Durham to serve as data for automated video and audio analysis and stimuli for perceptual experiments. This project represents just one example of an attempt to bridge the gap between ecological validity and experimental control that is often a barrier to overcoming the WEIRD problem, and I hope and anticipate that future music psychological research will continue to find creative ways to become less WEIRD.