What makes a great teacher? 

It's a simple question: what makes a great teacher?

Swathes of studies over the last two decades such as Hattie’s meta-analyses, Chetty, Friedman and Rockoff’s economic studies, reports by the OECD and the Rand Corporation all point to the teacher being the single most important external factor on student learning.

It is important to emphasise that the teacher effect is exogenous. Student motivation and student strategies for learning are more important than any external factor. As the Chinese philosopher Lao-Tzeu said: “When the student is ready, the teacher appears”. If something must be learnt, even a bad teacher will not stop it from being learnt. This much said, a good teacher can have a transformative effect on learning for reasons that are fairly obvious.

The next question, therefore, is what makes a good teacher? This is where complexity creeps in, where the endless uncontrolled variables affecting each subjective experience of learning muddy the waters. If the most rigorous form of research is the muscular and quasi-irrefutable experimental method of the double blind randomised controlled trial, then a big problem is that it cannot be used to measure what kind of teaching really has an impact.

In live classrooms, you cannot create serious control groups because each cohort will have a different teacher. But what is known as time-series experiments can be used (meaning that one tests several interventions on the same population group successively) but the natural growth and influence of external factors that are uncontrollable will affect learning. Qualitative methods such as surveys, focus groups or interviews can be run but, as anyone involved in qualitative research method knows, extending and generalising these findings to other population groups is contentious.

Most schools resort to student surveys since these are easy to administer. However, carefully conducted research shows that student surveys are highly problematic: they cannot be properly screened for a host of biases and contain very high error rates.

Some institutions will look at the results of massive globalised tests such as PISA, PIRLS and TIMSS, latch on to one or two top scorers (Finland and Singapore for example), seek some generalised account of what teachers (and systems) do in those countries and try to duplicate these practices in their own learning ecosystems.

There are two problems with this approach. First of all, ranking systems based on test scores need to be understood statistically (which is complex: how statistically significant are PISA score intervals between the first and the fourth placed, for example, and who can understand that?). Second, it is dangerous to try and apply a set of practices in one country with its history, demographics, culture, sociology, political framework, economy, languages and curriculum, to an entirely different setting.

Since educational leaders and board members are not always necessarily well-versed in educational research methods and since there is huge pressure on education systems to deliver, efforts to nurture great teachers can turn into a nightmare of absurd policies.

For example, grasping at bits and pieces of research, schools can start to insist that every teacher start lessons the exact same way, stop grading students, cancel homework, or talk less in the classroom. The clause 'the research shows that' will be thrown around eagerly, although rarely has the research been read from cover to cover and, when it has, insufficient degrees of confidence in findings and then need for further study are almost always prevalent.

Unfortunately, some myths in education persist despite fairly consequential debunking: such as the Mozart effect, Brain Gym, learning styles, multiple intelligence theory and a series of erroneous literacy strategies such as the 'three queuing' reading technique.

Would the teachers working in schools relying on these theories be considered 'great' for implementing such strategies? Here is the real problem: teaching only means something within its defined context. This goes beyond a simplistic judgement of institutions’ positions on what good learning and teaching is. It begs the question of whether one can truly speak of great teaching in such broad brushstrokes that the same behaviours and habits are expected in every context.

The sports coach, the PhD supervisor, the online undergraduate programme lecturer, the beginner’s language teacher, the teacher interacting with students with severe learning difficulties, the personal trainer, the grandmother telling a story, the highly methodical and conscientious teacher with low voice projection and bad handwriting, the charismatic, larger than life figure who nonetheless forgets certain tasks, the teacher who has a discussion with students that has nothing to do with whatever test is being run to decide where there is value added, each of these teachers can be a great teacher in very different ways.

In fact, what are known as epistemic practices suggest that learning and teaching are subject dependent; that seeking generalisable, transferable pedagogic practices is problematic because learning and teaching in chemistry is not the same as learning and teaching in visual arts or Chinese language instruction.

Because of all of these complexities, numerous experiments do not deliver. For example, The Gates Foundation poured hundreds of millions of dollars into a multi-layered experiment designed to boost teaching effectiveness through traditional approaches such as classroom observations, careful recruitment and performance bonuses but found that these created no significant gains in student learning.

Many efforts to answer this have provided results that make sense and seem reliable. The Sutton Trust’s 2014 analysis of 200 pieces of research identifies teachers' subject knowledge, understanding of common misconceptions, use of questions and assessment as factors that lead to excellent teaching whereas the Global School Leaders group shows how important it is for teachers to believe in their students.

Hattie’s meta-analyses, involving millions of data points, describe what he calls collective teacher efficacy as the most fundamental driver for student improvement. This means a collective belief, by a group of teachers in a school, that together they can improve student learning. Nonetheless, it should be noted that many argue that the quantitative research that is used by Hattie to determine good teaching is flawed.

These findings are not radical or counter-intuitive: who will argue that believing in students, the teacher knowing the subject matter or assessing well will not improve learning? Can we and should we extrapolate further?

The whole question of culture is worth investigating. To what extent is the transactional language of 'attainment', 'value-added', 'results' and 'scores' not so much an expression of scientific truth but of cultural bias? When administrators come across different teaching styles, different accents, different ways of communicating - and these are cultural - how do they feature on an observation sheet? And which evidence from which countries involving which learners was used to determine the checklist in the first place?

Qualitative studies among Aboriginal and Torres Strait communities in Australia make it clear that culturally responsive pedagogy is of vital importance to students. Cultural aspects such as reverence for the land and the wisdom of elders might not fit into the Western constructivist paradigm, and why should it? A study carried out on Chinese students showed that the ethical framework of the teacher was of paramount importance to them, something that had not been mentioned in teacher appraisal documentation. Students in South Africa, meanwhile, described the most important teaching factor as the teacher’s ability to engage in supportive dialogue with students.

School leaders have to be culturally responsive, taking into account the cultures of students and teachers before jumping at standardised descriptions of what good teaching is.

At my school, we ask teachers to set their own goals. We trust them as professionals to identify something meaningful to them and their ecosystem of learning. Certain coaching questions might be asked to dig deeper into the value and identity of those goals, but the goals are not prescribed exogenously, they come from within. This is because great teaching is not merely a predictable, observable, externally measurable fact as the social sciences would like it to be. There are too many nuances, shades of grey, areas of context.

This does not mean that we do not have exigencies and standards: feedback must be given, everybody should be treated with respect, care must be given for every child. However, the exact way in which that happens cannot be formulated for every classroom and every subject.

As schools evolve into more culturally responsive institutions, may they resist the urge to oversimplify what makes a great teacher and open their minds to different paradigms, the fact that teaching is not always the same and that context matters.