Yet another psychological debate with endless theories from either side. Nature Vs Nurture, Reductionism Vs Holism to name a couple, are joined by Qualitative Vs Quantitative in the major debates that we as psychologists are forced to ponder over till are heads hurt. I could name the pros and cons of each of the measures, but since this is such a repetitive argument, let’s get right down to the crux of the argument, the cons.
The main problem with qualitative measure that they are open to the investigators interpretation. Take for example if we were observing depressed behaviours within patients in semi-structured interviews. How do we actually define a construct such as depression? Sure, there may be guidelines and definitions to make diagnosis easier, but at the end of the day it’s up to the investigator or clinician in question to decide what are depressive behaviours. Even with inter-rater reliability (the whole point of these guidelines) behaviour will always be open for interpretation. This idea is supported by the work of Perez-Stable: ( http://archinte.ama-assn.org/cgi/content/abstract/150/5/1083 ) this study shows just a representation of the mis-diagnosis of depression within a clinical setting. Throw in other behavioural traits such as anxiety and obsession and were already encountering numerous reliability problems just based on diagnosis, let alone whether the patient is truthfully responding to conditions.
Well on the flip side we could use quantitative research method to try and find an un-bias measure of depression. Take the GDS developed by Sheikh and Yesavage: (http://psycnet.apa.org/psycinfo/1988-34658-001) a scale devised to identify depression within elderly patients. This scale would be admitted via self report measures where the patients would rate themselves on the scale. This in then turn would produce lots of quantitative results which could be analysed via t-tests and several measures of reliability, while at the same time reducing interpretation effects. Sounds amazing doesn’t it? There’s always two sides to every story and truth be told, quantitative measures are no exception. The problem with scales filled out by participants and patients is that they are able to show biases of their own, for example desirability bias. Participants will rate themselves how they see will be seen correct, beneficial, or “normal”. This effect is demonstrated by the work of Adams et al. (http://aje.oxfordjournals.org/content/161/4/389.short) Although this research is talking about desirability in amount of exercise, the effect can be generalised to desirability of depression and other conditions. Another problem with these self-report measures are leading questions: “how depressed would you rate yourself” for example. Even if the scale was from “not at all” to “very,” there’s an automatic assumption of depression within the participant. This may cause positive desirability in a lower than actual depression score, or negative desirability by over-rating how depressed they are.
I think the important thing to take note, is that regardless of many different perspectives of research there are, theories, or even ideas of human behaviour, there are always going to be infinite ideas of how to measure or support these ideas. With both quantitative and qualitative measures, there are most definitely faults, but if we combine both of them then we can eliminate or balance certain biases to some extent, and analyse our research to greater effect. Make the whole experiment and research more scientific, and isn’t that the point of psychology? Determine human behaviour in a scientific way? If we as psychologists can at least agree that both these types of measures have their place in research, then surely were half way there?