The Research Files Episode 107: Professor Alex Bowers on types of school leadership and targeting professional development

Hello and thanks for tuning in to this podcast from Teacher Magazine, the resource for K-12 educators published by ACER, the Australian Council for Educational Research. I'm Dominique Beech.

In today’s episode, I’m delighted to bring you my conversation with Professor Alex Bowers who joins us all the way from New York. He works to help school leaders use the data that they already collect in more effective ways, and his research focuses on the intersection of effective school and district leadership, data-driven decision making, student grades and test scores, and more. Our conversation today delves into some key pieces of research over the years that focus on school leadership and teacher professional development. He has some fascinating insights to share on what he and his colleagues have uncovered about different types of principals across the United States, and the alignment of perceptions between teachers and principals when it comes to leadership in the school. He also shares his innovative idea for re-organising how we think about teacher professional development. There’s a lot to cover, so let’s jump straight in. 

Dominique Beech: Professor Alex Bowers, thank you so much for joining Teacher. You're very much in demand around the world so I'm delighted that you were able to join us all the way from New York. You are a Professor of Education Leadership at Teachers College at Columbia University – to kick us off I'd love to hear a little bit about your background, your current role at Teachers College and also your main areas of research.

Professor Alex Bowers: Yeah, thanks so much for having me on. So, the kinds of work I do here at Teachers College – I'm a Professor of Education Leadership and I like to say that I help school leaders use data more effectively. And I mean that as far as the data conversation on both sides of the data equation in schools. So, on one side it's: How do we use the data that we already collect in schools more effectively to help build collaborative conversations, facilitate evidence-based improvement? How do we, as a faculty in our schools, bring evidence forward in a way that builds dialogue, trust and a shared understanding of the direction that we, as a faculty in our school, want to go based on what we know about our students, our conversations with our families? 

But then I also like to say that I work on the other side of the data equation. I do a lot of data science, a lot of data dashboards, predictive analytics, and thinking about helping school leaders find and use data visualisations and dashboards that actually address their evidence-based questions in their schools. Rather than … these days we have a lot of dashboards and data vendors who will create data dashboards for schools, never talk to educators, and plot them down on top of schools and everybody's kind of confused about why no one clicks on anything. And, you know, I take from my perspective that maybe we should include educators at the beginning of the data visualisation process. 

DB: I'm sure, hearing that, many of our listeners are very excited to hear more about your work – and we will be getting into some of your key pieces of research shortly – but first of all, obviously, as we've said you're based in New York and the majority of your research work has been in relation to principals and teachers in the US. Obviously, we're based in Australia, many of our listeners are based in Australia. I'd love to hear from you about what does that mean in terms of interpreting the data and the insights that could inform policy and practice in other education systems and countries?

AB: Yeah. The Australian context, I think, is absolutely fascinating. And I was lucky enough to be in Australia for most of February. And New South Wales, for example, is as you know the largest schooling system in the southern hemisphere. And, but in doing education research and practice and working with practitioners in New York City here in North America, you know New York City is the largest education system/school district (we like to talk about school districts in the United States), it’s largest in the United States. And I see so many parallels with New South Wales, for example, but also, you know, schools in Victoria, all across Australia, in that there's a lot of … One of the things I really was fascinated with in Australia is how far ahead of us you are in the neo-liberal marketisation of schools. The United States is headed in your direction in a lot of ways; I think there's a lot to learn about policy from just how Australia has organised schools and gets the work done.

At the same time there's a lot of the same structures; principals, schools, there's – well, I think we'll talk about it in a few minutes – we have a very similar focus on test scores and outcomes, assessments across … you know, English, reading and writing, as well as mathematics. The structures are very similar. And the kinds of work that I do in the United States is very much about, as we were saying, helping school leaders find the way to not tell teachers what to do. Especially with data – because, I think in both systems, teachers feel like data is something that's been done to them. The accountability focus and the test-based focus can be very problematic and unhelpful, I think, in a lot of situations.

And so, a lot of the success I've seen in the US system, but also in Australia, is when educators (especially education administrators and school heads) really see their role as facilitating a data conversation in which we’re able to bring together lots of different data far beyond test scores. So, like, school climate, academic climate, trust. How teachers trust, not only each other, but collectively how do we get along with the principal, the school heads? But also, how do we work together? And the kinds of conversations that we can have to build towards the kinds of success that we're looking for together in our schools.

I think there's a lot of parallels where we can learn from each other. 

DB: And so, I mentioned at the top of the episode we're going to be talking today mainly about school leadership and teacher professional development and the work that you've done in those areas. Let's start with school leadership first. You've done some fascinating research on the different types of principals in the US (Urick & Bowers, 2014). As you mentioned in the paper, it's a lot more complex than just simply a leadership style of a school leader, isn't it? 

AB: Yeah. The kinds of work that we've done – and much of this work is with a previous graduate student of mine Angela Urick (who is now a Professor at Baylor University here in the United States) – and, you know, this work is about the difference between transformational leadership and shared instructional leadership. 

So, traditionally in education and education leadership research, transformation leadership is seen as the principal is helping schools and the teachers really find their way together to become professionals around studying your own subject, bringing professional development, helping teachers find this idealised environment where they feel engaged and activated in their own professional development; where they can come together as teachers to learn together and build resources for the school and see their work as being authentic and engaged in the process. Where, instructional leadership traditionally has been seen as the principal leading the vision and mission of the school, leading the curriculum and then leading the professional development.

And in our work, one of the things that we've seen is that when we survey teachers on both transformational leadership and shared instructional leadership – and interestingly enough, the original work on this complex, multidimensional way of thinking about it comes from my advisor in my doctoral program, Susan Printy, so it's a long line of research which is fun – but it's this idea that they both exist at the same time. And, if you imagine a scatter plot of your school, and if you survey the teachers on both transformational leadership (maybe on the X axis), and shared instructional leadership (on the Y axis) – one of the things that's very interesting is, while if you then divide that into quadrants. So you can have low/low – so schools that don't say we don't have any transfer not doing the transformational leadership thing and we don't have any shared instructional leadership; or high/low – so, ‘we feel the transformation leadership is working, I feel empowered, I'm going out and getting resources I need, I feel like I could ask for things I get them, but I'm not really seeing the vision and mission and leadership of the professional development and the curriculum’ (so that would be high transformational leadership, but low instructional leadership); and then high/high – and these are schools that usually perform very well, when they have both.

What we actually don't see – and this was originally piloted with a small set of schools and then we replicated it nationally in the United States – there are almost no schools in the low transformational leadership quadrant but high instructional leadership. Which, from a theory level, would indicate that transformational leadership may be necessary, but insufficient, for shared instructional leadership. 

And there's been a really fantastic study from Paula Kwan in Hong Kong recently (Kwan, 2019) where she actually did exactly this study, where she surveyed a large number of schools about their transformational leadership and shared instructional leadership, but she also had the outcomes. And what she basically shows – she doesn't talk about it like this, but here's my analogy … it's like you have 2 dials. And, I don't know, in the United States our stovetop ranges, the dials are usually clicky, they click, you can click around, right, so you can really feel it. So, think of 2 dials, one transformational leadership and one shared instructional, in a school. And so, if you click the transformational leadership from low to medium (according to Kwan’s study) you can spin the instructional leadership dial all you want and nothing will happen. And the same is true if transformational leadership is low, right. If everything is low you can spin that instructional leadership dial all you want, nothing will happen.

You can work on vision, you can work on mission, you can work on the PD – like, you need to do transformational leadership first. Because once you click that transformational leadership dial to high, now the instructional leadership dial engages, and you can go from low to medium to high, and it actually makes a difference. But until you have high transformational leadership where the teachers feel empowered to build their own professional development together around their subject knowledge and as teachers in their schools, the work around shared instructional leadership really doesn't get you very far. And so, it seems to be that there's a balance between the 2; you want both at the same time. 

DB: A quick note to listeners that I will link all of these papers that we're discussing today in the transcript of our podcast episode that you can find at teachermagazine.com, because it's absolutely worthwhile accessing the full papers to read more in depth about this work of yours. The next study I wanted to touch on, Alex, is one that draws on data from TALIS (Bowers, 2020). Our teacher audience will be familiar with TALIS, of course that's the OECD’s Teaching and Learning International Survey and it provides some rich data to work with from lower secondary schools. You did some analysis around distributed leadership, which explored joint teacher-principal perceptions. Can you tell me about that?

AB: Yeah. One of the things that's fantastic about TALIS is that it's – not the most recent administration, so the 2025 just came out which is really exciting; there’s so much data to look at – and thank you to all of the teachers in Australia who filled out the survey, I know it's a lot! But my previous work, I was very lucky to have a fellowship with the OECD where I got early access to the 2018-2019 TALIS administration, which at the time included 157,000 teachers across 57 countries, 9,000 schools. 

And so, one of the things I wanted to understand was an idea that I've called a congruency typology model of school leadership. And what that means is, first, to what extent do the teachers agree with the school leaders about how it's going in the school? About the leadership across the kinds of items that TALIS asks. So, just like we were talking about with transformational and instructional leadership, the TALIS survey asks questions about these kinds of concepts. So, it's very similar in distributed leadership as well. And so, I was really interested in the congruency, this alignment between teachers and principals.

Because if you imagine, again, 2 axes – an X axis in a Y axis. And teachers could be overall thinking about how the school leadership in the school is going, how they feel about it, they can be ‘we don't do these things’ (low) or they could be high. But the principals could also be low (‘we don't do these things, I know that we don't do these things’) or high, right? And so, you can think of 4 quadrants where 2 of them are congruent/aligned. The teachers and principals could say ‘we're high’ in all these items in TALIS and they're aligned, but they could also be aligned where we agree we don't do these things, and it's a problem. 

So, one of the interesting points is in the study, first I did that congruency to be able to ask across these 2 levels (teachers and principals), to what extent across the countries do they agree on principal leadership and how it's going? And, interestingly enough, in that 4-quadrant idea of low/low; low/high; high/low; and high/high I only found 3 of the 4. I don't really see principals being way below what teachers say they're doing. Either you have congruency where we all agree ‘we don't do these things’, or we all agree ‘we're great, it's going great, we're doing all the things’. Or principals are high and teachers are low. And you can hypothesise that either the principals – we have the phrase in the US, ‘pie in the sky’ – like they think that they’re way ahead or above, and maybe they're out of touch with their teachers or maybe they’re aspirational. Or maybe the teachers know what they're talking about and it's the principal who's out of touch. 

But then what was interesting is I continued the work to be able to understand this rich complexity of the differences between how teachers and principals think about school leadership and then I faceted it by national context, and what I found shocked me. I found 3 different kinds of school leadership contexts across this almost 50 national contexts in TALIS. And what was really, really interesting is that both the United States and Australia only had 2 of the 3 types.

DB: Wow, that's really interesting. And so, we're exploring that area of perceptions – what principals and teachers think about what they perceive to be happening at the school, like you've just said in relation to that TALIS data there. The final piece of research that I want to talk about really gets us to think about the teacher professional development and what that might look like. It's really interesting to think about whether we can group or cluster different school types so that they can have similar PD that's more relevant to them. Again, fascinating piece of research that you've done here – there were lots of factors and indicators feeding into this. You looked at teacher perceptions of 5 things happening in their school – and again this is from survey data over in the US (New York schools) (Duff & Bowers, 2022). Can you tell me a little bit about what you found here?

AB: Yeah, of course. The survey that we used in New York City is called the Chicago 5Essentials survey and it's from a book by Anthony Bryk, published in 2010, about Chicago which is another large-sized urban context or city in the United States. It's a climate survey of the school to be given to teachers. And it asks questions about effective leadership, collaboration, trust, academic rigour, and a supportive environment. And the survey is open and any school can actually use it (it's in English). 

And what was fascinating was that, in New York City teachers every year take this survey and then the data is all averaged together for the schools. So, it's anonymous at the teacher level, but it's reported publicly for each school. So, in the study – and this work was in collaboration with Megan Duff who is now a Professor in North Carolina here in the United States. What was really interesting is we wanted to know to what extent is there one, or more than one, type of school responder to this 5Essential survey? Across the items, do schools generally respond the same way? Or are there like different patterns of responses? And do they statistically cluster together? 

And so, this Latent Class Analysis modelling is really cool in that you can do exactly that. You can ask, are there specific kinds of clusters of significantly different types of responders to your survey? And the analysis framework actually comes from marketing originally, which is a little problematic in education. But, you know, in marketing they segment their markets all the time to find out, you know, why you're buying this for different reasons. And we can use the same kinds of analytics. It's also a very similar algorithm to Netflix or Amazon – you watch this movie, these are the other 5 kinds of movies you might like, right. So, your teachers responded on the climate survey this way, here's 5 other schools who are most similar to you, right. It's almost the same algorithm.

And so, what we did is we had over 1,500 schools in New York at the elementary and middle school level (so this is all like basically primary school) and we had all this data publicly at the school level. So, we applied this analysis technique, and we found 6 different types of school climate in New York City. So, let me just talk them through.

At the top we had the versatile type. They were about one-third of all schools, and they are at the top of the scale. They say: ‘we do all the things, we do effective leadership, collaboration, trust, academic rigour in a supportive environment. We’re doing these things, we’re amazing, we’re an awesome school’.

Then the opposite were what we called the demoralised schools. They were 19% of all the schools in the city. And what's cool, what's nice about the analysis – sorry, I'm also a bit of a methodologist, so I kind of love the methods – but it's interesting, because you get these ratios out of your number of schools (this is every elementary school in New York City) – so the ‘demoralised’ are about a fifth of all the schools. And they are actually at the bottom of the scale, they answer the lowest. And one of the things, though, I say – really to anyone, but especially administrators when I present this work – is, I like to say that I believe that the ‘demoralised’ teachers and schools are probably the most important teachers in New York City. Because they're willing to tell the truth, right. They’re saying: ‘We have challenges in our context, things are not going like I want as a teacher. We need help – come help us’. And so, I really value their perspective.

And naming these groups is actually one of the hardest parts of the analysis, because you never want to blame people. It's a system that's the problem, right. The teachers are fantastic people and work incredibly hard and, you know, we need to cherish our teachers. And so, this is all to say that the ‘demoralised’ group, I think they're pointing to the kinds of resources that they really need. 

And then I'll go up from ‘demoralised’. So, then we had the controlled group. We called them ‘controlled’ because they're one of the highest on saying they have instructional leadership, but then for collective responsibility they’re low. as well as they’re a little lower on teacher-teacher trust, but they’re higher on teacher-leader trust. But they're low on academic press and they don't talk about individual students very often. So, OK, just keep that in your head. 

But then going up, we have responsive. We call them ‘responsive’ because they're actually typified by saying: ‘we don't have any leadership here’. They’re actually the opposite of controlled. It's all teachers. Their teacher-teacher trust is very high; their teacher-leader trust is very low. And they actually talk about student/individual issues quite a bit.

Then we had the developing. And they're kind of on their way to the ‘versatile’; they're going up the scale. And then we had collaborative, which is about 13% of all the schools. But I was really interested in ‘collaborative’, because if you just look at effective leadership, collaboration and trust, they look just like the top schools (the ‘versatile’ schools). But when you add in academic rigour and supportive environment, the collaborative schools look like the bottom schools. They don't have very much academic press (so, how hard you press the students academically) and they don't talk about student/individual issues very often.

And when we then added demographics to the model to see who these schools serve, then it's the ‘collaborative’ and then ‘controlled’ and ‘demoralised’(the controlled/demoralised had some of the lowest ratings). They serve the highest need schools, so they have the most minoritised population and highest percent of teachers with less than 3 years of experience. 

And then we looked at the actual test scores for English and then Mathematics … So ‘versatile’, at the top of the scale, they had the highest test scores by far they’re right, they’re the best schools in the city. What was interesting is that the ‘controlled’ (who had high leadership, but it looks a little authoritarian), the ‘controlled’ and the ‘demoralised’, we couldn't tell the difference in how low their scores were. They're both at the bottom – which is highly problematic for a Professor of Education Leadership, by the way – which we could talk about more. But then, the ‘responsive’ actually were kind of in the middle and the ‘developing’ were on the way up, like I said. 

But strangely enough – and one of our biggest and most surprising points – was that the ‘collaborative’ schools, who serve some of the most minoritised students and kind of look like the top ‘versatile’ schools when it comes to effective leadership, collaboration and trust – but they say they have very low academic press, and they rarely talk about individual student issues; they had some of the lowest test scores. They were almost as low as the ‘controlled’ and ‘demoralised’.

And so, this like really comes down to this idea of thinking about how we can think about using data. Like, what the teachers actually say, to be able to help understand – and not blame – but help understand schools and their contexts from a deeper perspective using the data that teachers actually provide. 

DB: And one of the most interesting things for me from this particular study of yours was the link to student outcomes, and you've mentioned there, you know, the correlation with test scores for students. Was there anything else to mention on what you were finding for the link to student outcomes? 

AB: Well, it's just so interesting that we have – across both the Australian and US context – we're always talking about the test scores. And teachers though are right to say, ‘but that was last year’ and ‘we've got the issues we have today’. 

And so, I think what’s super helpful to improvement discussions in schools around data is to include both the outcomes (the test scores) and these other kinds of data like the academic climate, trust, collaboration, academic press, how often we talk about individual student issues. And notice that this is also systems level data. I think in a lot of ways when it comes to thinking about outcomes and other kinds of data that we collect in schools, as educators we can get really focused on students. Like a lot of times for some of the systems-level data we’ll open the datasets in collaboration with educators and, you know, the data are public many times and so, of course there would be – you don't have individual student data in public data, they’re averages for the schools – but educators will always ask ‘where's the student data? I need to see the individual students’. But what's interesting here is being able to see the systems level and to be able to compare across schools to be able to understand which school contexts are similar in our school system that could maybe speak to each other? And not on test scores, but on how we actually do the work and think about our approaches to education.

DB: Really fascinating. It offers some real food for thought on when we're thinking about the sort of PD that could be delivered to schools and how we could – as I've said before – cluster them to better meet them at their point of need. So, you're really grouping ones together that are facing similar challenges, aren't you? 

AB: Yeah. And one of the things to think about – I think it's true in Australia as well as it is in the United States – is that school level professional development, I have a hypothesis that school level professional development … and we only have 2 kinds. We have one-size-fits-all or completely individualised, neither of which work. And so, my innovation is to say, what if we listen to teachers, survey them on the very specific issues that they have, that they report on their practices around climate – like the 5Essentials survey? And if we then cluster schools together based on the similarities in what the teachers say the successes and challenges are around these issues, we wouldn't have one-size-fits-all, or completely individualised, maybe we have 6 types.

And so, for example, to spend professional development money on the same intervention for the ‘versatile’ schools (who are at the top of the scale) and the ‘demoralised’ schools – to give them the same intervention, you might as well have a pizza party with the money because it'll have the same effect. Well, it actually might be better. 

And so, what if instead, we use these kinds of pattern analytics with the data that we already probably collect across most of our schools in many different ways (or could easily) and hear and listen to the teachers about what they say these challenges and successes are? And then bring schools together who are similar. And then we can target the professional development to those specific needs. 

Think about a heartbeat, even, or different ways of personalising the professional development to the very specific kinds of issues such that then we can learn together around our similar issues, we can share. Or you can partner diatically with other schools in different clusters to understand their different contexts. But we can build resources around the actual problems that teachers say we have in our system and we get a lot of benefits from this kind of idea. Especially if the survey stays anonymous at the teacher level, then the survey is valued, teacher voice is valued, and people will buy in more to the professional development that's created and offered and have a voice. And the hope would be, see success in the system.

And to go back to one of the things that we were talking about earlier, so I said that Australia and the United States have 2 of the kinds of leadership globally, not all 3. And so, there was a very nice study a few years ago (Bowers, 2020) where they showed that – across the OECD countries who all take TALIS – there's these different contexts of school leadership. And they basically clustered the United States and Australia together as well around this test-focused construct. And it's the third type of leadership, that we do not have between our 2 systems, is a kind of leadership where the walls of the school begin to break down in that teachers feel like they’re professionals in their work and can move between schools in a regional network of professionals learning together. 

And so, this is my goal with this kind of work – to use the data that we have and the data analytics that we can pattern together, to bring teachers and schools together across the walls of the schools. So that we can learn from each other’s organisations and not feel is isolated like we currently do in our own schools and be able to share our innovations around our common problems.

DB: Wow, I'm sure there are many people listening to this episode who were enthusiastically nodding their heads as they listened to you speak, just like I was. That was such a delight, Professor Alex Bowers, to be able to listen to you share your insights and expertise with our Teacher listeners. Thank you so much for taking the time today.

AB: Thanks so much for the opportunity. It was great. 

That's all for this episode. Thanks for listening. Don’t forget to visit our website, teachermagazine.com, to explore the references and related reading at the bottom of the transcript for this podcast episode to read further into Professor Alex Bowers’ publications, because our conversation today really only scraped the surface. You’ll be able to find it all under the podcast tab. If you enjoyed this episode, please take a quick moment to follow our show on your podcast app, if you haven't already, and leave us a review. Both of those things help more people like you to find our podcast, and they're a really big help to our team. We'll be back with a brand-new episode very soon.

Teacher magazine is published by the Australian Council for Educational Research.

References and related reading

Bowers, A.J. (2020) Examining a Congruency-Typology Model of Leadership for Learning using Two-Level Latent Class Analysis with TALIS 2018. Organisation for Economic Co-Operation and Development (OECD) Publishing: Paris, France. https://doi.org/10.1787/c963073b-en

Bryk, A. S., Sebring, P. B., Allensworth, E., Luppsecu, S., & Easton, J. Q. (2010). Organizing schools for improvement: Lessons from Chicago. University of Chicago Press. Organizing Schools for Improvement: Lessons from Chicago, Bryk, Sebring, Allensworth 

Duff, M. & Bowers, A. J. (2022). Identifying a Typology of New York City Schools Through Teacher Perceptions of Organizational Capacity: A Latent Class Analysis. Leadership and Policy in Schools, 21(4), 791–815. https://doi.org/10.1080/15700763.2020.1854789

Kwan, P. (2019). Is Transformational Leadership Theory Passé? Revisiting the Integrative Effect of Instructional Leadership and Transformational Leadership on Student Outcomes. Educational Administration Quarterly, 56(2), 321-349. https://doi.org/10.1177/0013161X19861137

Teachers College Columbia University. (n.d.). Alex J Bowers Faculty Profile. https://www.tc.columbia.edu/faculty/ab3764/

Urick, A. & Bowers, A. J. (2014). What Are the Different Types of Principals Across The U.S.? A Latent Class Analysis of Principal Perception of Leadership Styles. Educational Administration Quarterly, 50(1), 96-134. doi:10.1177/0013161X13489019.