The approach

Transcript

James Ladwig: “Two sides of this, two basic tasks, the first part is taking a look at the observational data itself and kind of exploring that. The basic question in my head when I was looking at that is, ‘Okay, what does the data say about the various elements of pedagogy and how they fit together.’ You know, you’ve probably heard others talk at length about how pedagogy works and have all kinds of ideas, well the question is you know, ‘Does the data actually confirm some of these? How does it, how does it work together?’ The goal of looking at that was to see whether or not I could use that data to come up with a measure of the quality of pedagogy that I would then look at in relation to outcomes. And so that’s the second part of what I’ll, ah talk about. Once you get to that far then the next question is okay can you link that in some way to the outcomes of the students, that’s a lot more complicated than it sounds, particularly in a secondary setting when you have kids in multiple classrooms with multiple teachers that change every time. It poses a number of problems, can I come up with a shortcut, just to see whether or not it’s worth the longer-term effort of doing a more complicated one? Um, so I’ve got the shortcut version here. Ah, the outcome data we’re going to keep in mind, I’m reliant on what is available from asTTle data so at the time I was doing this, the rest of the team is busily trying to get all that data compiled and so we picked the year where we had the biggest number of students with comparable data. So when I said we were reliant on our data warts and all, it’s a combination of, you know, getting it all in, in an order where you can actually use it, okay there are some limited options I want to talk through some of those. And the last is give you some of my interpretations about the implications, for things schools and teachers and Te Kotahitanga can think about. You’ve got the issue of whether or not the pedagogy’s been culturally appropriate, whether it’s been responsive, culturally responsive, there’s a sense of caring. Whether, whether or not the lessons were well managed and high expectations for performance and for behaviour, particularly from our students. And in a sense that gives you a kind of two dimensional model, alright.

“So the first task was to see whether or not we could model those statistically. Now they’re different kinds of information so I had to do different ways. The things that are counts, you know you’re counting what kind of feedback there is, the observations of the five kids, five students, whether or not they’re engaged, are very direct counts, they, they’re called direct observation measures. The cognitive level, um, is kind of indirect because it’s a person at the end of a lesson making a judgment overall, a kind of global judgment but it’s only one indicator, so that’s, um, a kind of separate thing.

“On the back though the caring relations issues are basically various ways of getting at a larger concept, or an underlying concept that you can’t look at directly, right. It’s what, statisticians would call a Layton construct, it’s underneath. And in statistical terms there are ways to model, that is see if the data fits this bigger thing. One thing you may want to know is the six items of the caring relations are all correlated which means if one's higher the others higher, if one's lower the other's lower, and, you know, to various constructs. That being the case it told me it was a good shot it would work as a, a Layton construct and then when I went to do that, this is what the picture will look like. Does it make a good construct, the answer is yes, with some slight modifications and it is I thought you might find the modification interesting. That is the issue of a well managed classroom, I actually don’t need in the measure, it doesn't add anything more, I, it’s a better measure without it, alright. And it doesn’t mean it’s not important to do and it doesn’t mean you don’t want a part part of the, ah professional development, but if you ever see some publications with the measure it won’t be in there, that’s why, alright. A whole bunch of statistical measures to tell you whether or not the data fits this model and the answer is yes. So you’ve got the caring relations inside this big idea and the, and the little codes on the back of the coding sheet, they work. It’s not only an idea, you can measure it, okay. So for success number one.”