In which I disagree with the Education Research Establishment

Education research is, largely, the laughing stock of the academic community, and I'm sorry to say that we deserve it. For some reason, it was decided that education research should not follow the model of the natural sciences, nor should it prescribe to the model of the social sciences.
Failing to adhere to a natural or physical science research model makes sense, in some ways. Educators are not studying purely natural phenomena, nor are tightly designed lab experiments feasible or particularly meaningful in determining the efficacy of educational practices. There is something to be said, however, for taking a leaf from ecology in our studies of large populations of students, if we were to treat those students as part of an ecological community that has its own social niches, other 'organisms' (teachers and administrators), and a complex set of interactions. However, this is not the most pragmatic approach when other techniques have been developed to study populations of human beings.



It is the rejection of the social science model that is most disturbing. Education researchers chose long ago, whether consciously or not, to separate themselves from psychologists and anthropologists, whose methods might be useful to the study of human behaviour and interactions. Furthermore, this separation occurred prior to the cognitive revolution in psychology, so much of what is (poorly) co-opted from psychology stems from the behaviourist period; education researchers who perform experiments that mimic psychology still treat interactions as stimulus-response, with little or no reference to what's actually occurring in the black box of the mind. Likewise, techniques borrowed from anthropology fail to encompass the whole of the student experience, along with other failings to be discussed below.
Firstly, education research would be well served to follow the protocols of cognitive and experimental psychology, rather than simply teaching some aspects of developmental psych to teacher-candidates and forgetting the rest of it. For example, Piaget's timeline of cognitive development is taught, complete with ages where each stage might occur, but we fail to accurately apply these when studying learning in children, or for that matter when planning curricula. This foolishness needs to end. More importantly, we as researchers could at the least be informed by, if not work alongside, the newer disciplines of cognitive psych and cognitive neuroscience, which are explicitly devoted to understanding how the mind/brain works, rather than simply the behavioural outcomes of an accepted or proposed teaching practice.
For illustration purposes, let's attempt a thought experiment: Suppose you have developed a revolutionary new teaching practice that stems from the observed manner in which younger children learn language; children appear to learn language simply from immersion in a culture that uses that language, and through exploring the possibilities. Before we attempt to apply this method to learning something else, sat maths, what would you need to know? Would it be appropriate to find out whether we actually learn language through immersion or whether something else is going on that simply looks like that? How about whether humans can, in fact, learn something other than language in the same manner? Is there a period of development for which this capacity, even for language, is switched on and after which it is dormant? What would the consequences in terms of performance (the cliche of student achievement) if students are not equipped with the cognitive skills to learn a new concept in that manner, or if the cognitive module that allows for exploratory learning is switched off or degraded after a certain age?
None of these have been addressed by the researchers who promote what is known as constructivism, or discovery learning. Rather, the relativist philosophy that we construct our understanding of the world through our experiences (and interactions with others, in the case of social constructivists) governs the prescribed teaching methods. This completely discounts any chance of the brain having hard-wired learning mechanisms, or that it might have different hard-wired mechanisms for different types of learning. The objection that will come from those who might have dabbled in neurobiology without grasping the finer details or examining the research is that the brain has a remarkable plasticity and rewires itself based on environmental factors. This is indeed true, but the rewiring is in response to new stimuli and not creating new modules, but building upon those that are already there.
Perhaps more damningly for education researchers is the current trend that leaves out even the possibility of making proper quantitative studies of learning feasible. The Journal of Research in Science Teaching, which is the flagship journal of science education in the U.S., shows a distinct bias towards qualitative, descriptive studies that have far too small a sample size to be applicable to any situation other than the specific one examined. For a practitioner-researcher who is conducting action research in their own classroom in order to tweak their teaching style to better suit their students, this would be acceptable, but it is a complete and total waste of time for a discipline-wide journal to publish this drivel. An interview with four inner-city Hispanic females about their views of science is exactly that: the views of four people on science. I can't use that information in any way other than to pad my CV if I happen to have written it.
Anthropology, being a study of cultures and groups of people, might be better suited to study situations that arise in schools other than learning. Want to study bullying, its causes, and ways to curb it? Well, apart from the psychology of power dynamics, it is clearly useful to understand the social situations that beget and abet bullying in an effort to diffuse those situations, thus reducing the practice. Unfortunately, education's version of anthropology is deeply rooted in ideas born of feminism and post-modernism, (critical ethnography, caring theory, and other such nonsense) and purposely abandons the methods of the social science due to what appears to be a protest of a white male dominated discipline. Seriously, folks, that's tantamount to committing academic suicide. We can't generalise from a data set of 10 people, no matter how descriptive you think you can be, nor is it useful to understand how you think white male society is keeping these kids down unless you can establish that they do in fact learn better in a group of nothing but those like them. If that were the case, you might be able to establish the point that it would be useful to re-segregate schools based on ethnicity and gender, provided (and I seriously doubt this last caveat would happen) we ensure somehow that what is provided is truly equal.
So how do we as educators fix this mess and actually gain the respect of the rest of the academic community? Treat our discipline as another social science: one that understands the value and use of statistical methods, one that welcomes the work of psychologists and neuroscientists who can tell us what is happening in our students' brains, and one that uses the methods of psychology and anthropology to understand the problems we all face, rather than pretending that we're an isolated discipline that has nothing to gain from outside influences. Let's dump the bizarre association with feminist philosophy that impedes our progress. This doesn't mean that we don't have to be sympathetic to the differences between genders and cultures in our teaching, but by emphasizing that we're all human together we can start to erode those deficits that do exist. Incidentally, apart from some aspects of spatial reasoning, those deficits do not appear to be biological in nature, so they can be overcome with earlier exposure to those areas where each gender or culture is deficient. There are fewer females than males in the sciences in large part because we as teachers fail to make it interesting to them, not due to any lack of ability or artificial means of keeping them out. Most importantly, it might just be possible to develop teaching methods that are informed by how the mind actually learns, and tailor them by subject area to those means, rather than pretending we have a one-size fits-all solution. We can drop all the pseudoscience around things like learning styles, stop claiming "Johnny/Mary can't learn this the way you want to teach it because of (insert bullshit sociological theory here)", and make some real forward progress on how to teach children effectively. That is, after all, why you claim to have pursued this profession. Stop asking "what will this publication do for my professional reputation," and start asking "what can we as a research community learn from this."