1
Why does psychology need methodology?
Peter Ayton
This chapter introduces the idea that psychology is a science that can and doesâand indeed, mustâpursue legitimate scientific methods to make progress. Unusually for a book of this kind some of it is written in the first person as it includes personal reminiscences; however the argument is not based on personal experience.
Aims
The aim of this chapter is to instil in the reader an understanding of the need for a scientific approach to the study of behaviour. The reader will be invited to consider the merits of the scientific approach as well as some of the challenges that this approach presents. After reading this chapter the reader should appreciate the need for scientific methodology in psychology as well as the potential for applying it.
Introduction
It is a fairly typical experience in the life of a psychology student to discover, shortly after they have embarked on the course, that it is not quite what they thought it might be. Although some new students have no really clear idea of what to expect (I had absolutely no idea), the image of human psychology is for many rather exciting, exotic, and mysterious. Somehow one expects to be dabbling in the psycheâusing secret methods to reveal the hidden nature of human beings. In my view thatâs exactly what the methodology of psychology does enable one to doânonetheless students often complain that laboratory and methodology courses are rather âdryâ and technical. When I was an undergraduate psychology student one of my contemporaries complained to our tutor that they had chosen to study psychology because they: â...wanted to understand peopleânot to measure them doing âsilly thingsâ and turn them into numbers...â. At the time I was rather impressed by this argument and rather shocked by my tutorâs characteristically robust response. He answered by suggesting that, if the student simply wanted to ponder the nature of people, then they might be better off reading the novels of Jane Austen or Tolstoyâbut we were here to do science.
This sharp remark echoes the distinction that was identified by C.P. Snow in the 1950s according to which there are two separate cultures of thought that are attributable to the nature of educational attitudes. The culture of the literary intellectuals and the culture of the scientists view each other with mutual suspicion. According to Snow (1959) the position was then worsening; he could recall there had been a time when the two sides could at least manage âa frozen smileâ but that they were by then incapable of communicating. My tutor and student colleague certainly didnât manage a smile of any description. Instead there was one of those awkward moments when two people discover that they have attitudes that will require more effort than they believe that the other is prepared (or able) to make to understand why they are wrong.
As an example of the difficulties that were produced by the separation between the two cultures, Snow pointed to the widespread ignorance of the second law of thermodynamics that was displayedânot just by the lay populace but by most contemporary intellectuals. It is as if laws of science are not considered relevant for understanding the world. As an illustrative example for psychology we could consider the widespread belief in astrology. It strikes me as indicative of a major problem somewhere that most people are apparently aware of the idea that there are twelve sorts of people whose personality is determined by their birth-dateâbut simultaneously quite unaware of any scientific theory of personality. Typically astrology proposes, for reasons that are never clear or even announced and with no reporting of evidence, that your behaviour and its consequences are determined by the movement of the planets. Meanwhile psychology struggles to specify and justify valid theories by testing them with carefully measured observations. Despite the self-critical nature of any science (scientists are, quite properly, forever criticising each otherâs theories) it seems that some people exclusively reserve their scepticism for scientific rather than unscientific claims. Belief in astrology as a means of predicting behaviour appears to be widespread; I have even met psychology students who are prepared to admit (often somewhat more defiantly than the term âadmitâ suggests) their belief in astrology! At the same time many students will wonder about the value of scientific approaches to the study of behaviour1.
Is psychology a science?
All this talk of science and its merits and status requires some explanation. What exactly is science? There is no big mystery here, despite the rather portentous tone of some of the discussions of the nature of scientific enquiry. My dictionary defines science as systematic and formulated knowledge or the activity by which this knowledge is pursued based on observation, experiment, and induction. Astrologers do not seek to obtain knowledge by conducting systematic experiments. Astrologers use their methods simply to pronounceâthey are not in the business of collecting observations to explore whether any theory they propose is false or in need of modification. As a consequence they cannot be considered as scientists. For the same reason we can also exclude graphologists, scientologists, and most psychoanalysts. But why shouldnât systematic knowledge about behaviour, including human behaviour, be pursuedâand even discovered?
It is a fairly common task for the introductory student to be asked to write an essay discussing whether psychology is a science. I once discussed this habit with an eminent professor of psychology who told me that it was a mistake to ask students to write such an essay. âIt would be the same as asking them to consider whether a chicken had lips,â he said. âYou have to consider what lips are, what chickens are and whether the former can be found with the latter. All this could be done of course and it could be a quite rigorous discussion, but to what end?â Perhaps such a debate (about psychology and scienceânot chickens and lips) is all very well for clarifying approaches to application of the scientific method in psychology but, plainly, psychology as practised is a science as it uses systematic methods for attempting to acquire knowledge.
Somehow, though, there is doubt that psychology really is a science. In Britain the main government body that funds psychological research has been required to change its name from the Social Science Research Council to the Economic and Social Research Councilâscience being neatly and quite deliberately excised from its heading. When the UK government had a Department of Education and Science (science was dropped from its brief recently) it did not consider psychology as a science subject at all, grouping it with such subjects as law and librarianship (Radford, 1982). Could these government perceptions be a symptom of Snowâs two cultures? The notion that it might be useful to be scientific in order to understand people is often viewed with suspicion. But what is the alternative?
The style of the early part of my science education invited the conclusion that all the worldâs mysteries had been largely solved by the end of the nineteenth century and my job as a student was simply to learn what had been discovered. When we conducted experiments in physics, chemistry, or biology they were usually in the nature of demonstrations of principles and effects that were considered well established truths. In a sense it was more like science history than science per se. All around the walls of my school science laboratory were pictures of eminent Victorian scientistsâusually men with long grey beardsâwho glowered down as we struggled with Bunsen burners and test-tubes in our efforts to demonstrate what was known to be true. Sometimes our experiments âdidnât workâ and we were left feeling incompetent (unless the failed experiment was conducted by the teacher in which case we were jubilant).
Never, as I recall, did we conduct an experiment that we had devised ourselves in order to test the validity of a new idea that might add to the existing stock of knowledge about the world. No doubt this is due, at least in part, to the fact that there was a lot to learn about the relationship between scientific theories and laboratory phenomena. Given the huge amount of scientific knowledge in the traditional sciences, simply learning about what was well established took years. Nonetheless I was left with the impression that science chiefly involved the application of well understood principles. Those grey-bearded faces on the walls seemed to be saying that they had solved all the worldâs mysteries and it was simply my job to learn what they had discovered. It never occurred to me until years later when I studied psychology that science was an ongoing and very much alive process of acquiring and testing new ideas.
In terms of creating the impression that all is solved, psychology is perhaps obliged to be a little more modest in its approach; for a science that is officially only a little over a century old there may be relatively rather little to serve up in the way of impressive demonstrations of well established truthsâalthough there are some. The very first experiments that students will carry out are usually conducted in order to reveal well established reliable phenomena. Before very long however, students will be prompted to devise their own experiments testing their own hypotheses. The nature of psychology invites people to challenge and extend the current state of knowledge. Whereas early laboratory experiments may consist of demonstrations, fairly soon the psychology student will, or should, feel encouraged to embark on their own train of thought and attempt to do something new. In order to do this, though, one does need to know how to go about acquiring knowledge.
Methods for acquiring knowledge
In his defence of empirical psychology Broadbent (1973) identified three possible methods for acquiring knowledge. First, one might assume that one could just wander about the world observing as you goâmore or less at randomâand the facts of nature will inevitably compel you to induce certain generalisations. This no doubt happens to some degree (perhaps this is where astrology came from) but it is not recommended as a method for obtaining valid theories. In the first place, unguided intuition is not always reliable; people can come to conclusions that would not be supported by an objective perusal of the facts. Enormous feelings of confidence about an explanation are no guarantee of its viabilityâas psychological research has established (cf. McClelland & Bolger, 1994). Furthermore, it is rather limited as a means of obtaining the observations in the first place. As the rest of this book teaches, it is desirable to structure oneâs approach to the measurement of phenomena in order to be able to explain them. In order to measure systematically it is often necessary to create particular conditions where measurements will be unaffected by interfering factors. It is this requirement to be systematic that results in the formulation of methodology for collecting useful data by, for example, conducting experiments that are specially designed and controlled so as to enable clear measures to be made. In order to gather observations selectively one has to adopt a strategy for doing so. This strategyâthe methodologyâwill vary according to the circumstances that one is dealing with, but, as this book teaches, there are certain tricks of the trade.
The second method identified by Broadbent is what is referred to as the hypothetico-deductive approach. This approach is systematic. Here one proposes a theory, predicts the consequences from it and organises the collection of observations selectively so as to determine the truth or falsity of the predictions from the theory. The hypothetico-deductive method is a popular model for psychological investigations. Most scientists do not wander about observing at random and they are also fond of pointing out that one can never prove the truth of a theory but only disprove it. Thus, if you make clear predictions from a theory you have specified a means by which it might be shown to be false. If the theory survives the empirical test then one can feel that much more confident in the truth of the theoryâor at least one has shown that, for the experimental situation studied, it works. It is easy to see this approach as a sort of competition for survival of the fittest. A theory that is testable will only survive if it successfully accounts for the observations used to test it. Thus good theories would survive scientific test and bad theories would either have to be adapted to pass the test or be discarded. By this method science could be seen as progressing on an evolutionary basis.
There is another way howeverâthe third method identified by Broadbent. What is known as the Bayesian approach2 advocates that one should not consider just a single theory but consider what alternatives there might be. We can think of it as a market-place of ideas with competition among a number of alternative theories. Advocates of the Bayesian approach have criticised simple hypothesis testing on the grounds that psychologists have sometimes tested their favourite theories by making predictions that could also be predicted by a number of other rival theories. If a number of different theories make the same prediction as the one that you have set out to test, then there will be no progress in determining the best theory. Moreover, if the prediction turns out to be inconsistent with the factsâi.e. wrongâthen you will be left literally at a loss to explain what has been observed.
When faced with this outcome I have seen students attempt to explain away the results of their experiments. For example, they will often argue that perhaps the subjects in their experiment didnât properly understand the instructions, or perhaps the room was too noisy and distracted them, or perhaps they werenât motivated to attend in the first place. Now all of these are legitimate possibilities but, of course, one should also at least consider the possibility that the theory that led to the predictions is simply wrong; perhaps people just donât behave in the manner implied by the theory. In fact if we are not prepared to consider this possibility then there is very little point in going to the trouble of conducting the experiment in the first place.
The idea that the scientific theory might be wrong can take some getting used to. After the shock of discovering what psychology is like and learning to cope with methodology there is for many students another far more traumatic shock. Psychologists are not like the men with grey beards in the picturesâthey are people who donât know all the answers! This can create a good deal of discomfort for students who have been taught to expect that science is full of answers and reliable truths. Sometimes, as a consequence, it will be tempting to dismiss psychology as not a true science. However, if one sees it as a challenge to contribute to the progress of the subject rather than be despondent about its (lack of) achievements to date, the result can be quite exhilarating. Something needs to be done and proper methods of enquiry must be pursued in order to do it.
Manyâs the time I have heard a student complain that their experiment didnât âworkâ. But in a sense all experiments âworkâ because everything always happens how it âshouldââhow could it not? However, the theories, or the predictions, or the way in which the predictions are measured and tested, may not be valid. I suspect that the basis for the âit didnât workâ complaint (particularly for its typically despondent tone) may be a hangover from the perception of laboratory science as consisting of a series of demonstrations of what is known to happen for reasons that are known to be true (those pictures of Victorian gentlemen scientists again come to mind). In fact the state of the science of psychology (and, in truth, the other sciences as well) is much more exciting than that. To a very large extent we simply donât know how things happenâthat is why we are engaged in science in the first placeâwe are trying to find out. Trying to find out why things happen as they do is not easyâin psychology, for reasons considered in the next section, it may be much harder. The process of conducting scientific psychology entails much critical questioning and rigorous examination of theory, method, and conclusion. But that is not a reason to abandon the effort to engage in scientific psychology; in fact it is a very good reason why psychology must apply scientific standards with even greater effort.
Some objections to experimental psychology
It cannot be denied that there have been many thoughtful and highly influential criticisms of experimental methods in psychologyâmany of them coming from psychologists themselves. In this section I will briefly consider two of these objections and what can be said or done to counter or appease them.
Scientific psychology cannot be applied t...