19 Quantitative/qualitative
It is useful to differentiate two main traditions of empirical research: quantitative research and qualitative research. Traditionally speaking, these two research traditions have often behaved as opposing camps, with individual researchers aligning themselves one way and barely recognising work carried out the other way. However, it is nowadays not uncommon for music psychology researchers to sit between the two camps, incorporating elements of both quantitative and qualitative research into the same research programme and even into the same study.
Quantitative research is identified by a particular emphasis on objectivity and generalisability.
Subjectivity is minimised through the application of mechanistic data collection processes, which in the context of psychology means for example administering closed-form questionnaires (where participants might answer questions by selecting numbers on a scale from 1 to 7) or behavioural tasks implemented on computers (where participants might be presented with various stimuli and respond using a keyboard and mouse). Most commonly the results from such studies will be represented numerically, for example as scores from different subscales of a questionnaire, or as reaction times from a perceptual task.
Generalisability is pursued by favouring experiments using large participant groups sampled from a well-defined population, and by using so-called inferential statistical methods to estimate the confidence with which the results from a given experiment can be generalised to future experiments. These methods tie closely with the methods used in non-human sciences, for example physics and chemistry.
The objectivity and generalisability of quantitative research comes at a cost. Quantitative data collection processes typically provide only a very restricted window onto an individual’s experience at a particular point in time; this is a result both of the drive for objectivity (which mandates mechanistic data collection procedures) and of the drive for generalisability (which mandates large participant groups, which in turn also require relatively mechanistic data collection procedures). A necessary consequence is that quantitative studies are generally limited to testing rather specific research questions or hypotheses. The researcher is therefore limited to finding things that they were already looking for, and that introduces its own kind of subjectivity and bias into the scientific process. A particular risk is that the researcher ends up assuming a theoretical framework that imposes values or assumptions derived perhaps from their own cultural or socioeconomic background that are fundamentally inconsistent with the underlying experience of the people being studied.
Qualitative research methods seek to address these concerns. Instead of highly mechanistic data collection processes that try to minimise the subjective role of the experimenter, qualitative studies include the experimenter as an integral participant in the data collection process. This might involve face-to-face interviews with participants (e.g. one-to-one interviews, or focus groups), but it alternatively might involve the analysis of free-form text collected either from questionnaires or from naturalistic corpora (e.g. newspapers, chatrooms, blogs). Crucially, the qualitative approaches recognise and value the experimenter’s role in responding to structure as it appears in the data, for example by extracting key themes from the interviews and coding the data in terms of these themes.
Involving the experimenter in this manner is necessarily time-consuming: for example, it takes much longer to interview someone one-on-one than it does to email them a questionnaire. A consequence is that qualitative studies often use much smaller participant groups than quantitative studies. In the context of quantitative research, small participant groups are generally seen as a fundamental limitation on account of the goal of generalisability: with a small participant group, it’s difficult to tell to be sure whether particular observed patterns are common across the population, or whether conversely they just reflect idiosyncrasises of the particular participants that happened to be recruited. In the context of qualitative research, small participant groups are seen as less of an issue, because qualitative research places lower value on the goal of desirability, and higher value on the goal of understanding the idiosyncratic experiences of a particular collection of individuals.
Music psychology has important roles for both quantitative and qualitative research. An key goal in music psychology is to understand the psychological mechanisms that underlie music perception and production both from a neuroscientific perspective (how do the neural substrates of the brain support these musical processes?) and from a cognitive perspective (how can these musical processes be understood in terms of information transmission and information processing?). Quantitative methods are essential for both these kinds of research. However, music is also fundamentally experiential, eliciting incredibly rich intuitions and personal experiences, and fundamentally social, being deeply embedded into many aspects of human societies across the world. Qualitative methods are invaluable for exploring this richness. Correspondingly, we will explore both quantitative and qualitative methods in this course.