Vocational Voices: Season 1, Episode 10
The impact of schools on young people's transition to university
Steve Davis (00:00)
Hello I'm Steve Davis. Welcome to this podcast for Australia's National Centre for Vocational Education Research.
Do you remember your school years? And do you remember them fondly or with some angst or embarrassment? Would you be surprised to learn that a number of factors relating to your school life played a critical role in guiding you towards or away from university? In the paper, The impact of schools on young people's transition to university, researchers have sifted through data from the Longitudinal Surveys of Australian Youth and discovered that while a young person's individual characteristics hold the greatest sway, certain school attributes account for a variation in a student's tertiary entrance rank of up to 20%.
I asked one of the paper's authors, Tom Karmel, from NCVER, to give us a brief overview of these longitudinal surveys and describe what characteristics of schools they were able to tease out from the data.
Tom Karmel (01:01)
The data come from the Longitudinal Survey of Australian Youth, which is a long standing longitudinal data set, which dates back quite a long time. The particular data in this study comes from the 2006 cohort. What was wonderful about this data set is it actually collects some data about the schools, and this is what we've tried to use in this particular study.
There are quite a few variables that it does collect. First, fairly obviously it collects the school sector where the school is a government school, a Catholic school or an independent school. It collects how big it is, so the number of students at the school. And then there are a quite a few variables that we actually create which reflect the characteristics of the school.
Steve Davis (01:52)
And from this set, which I believe, I mean something like 25 different factors or thereabouts that you were looking at. Can you give us an insight into what school characteristics were the most dominant in influencing students pursuing university education?
Tom Karmel (02:10)
Well, the first thing that I need to refer to is that in these models, we get a split between the importance of school level effects and the importance of individual effects. And at the end of the day, it is the individual effects that dominate the characteristics of the individual student. And that explains around about 80% of the variation in the TER scores.
But nevertheless, the school level factors explain about 20%. We can split that 20 percentage points into two components. One is the proportion that is explained by our variables and the other is something else, that is something we measure statistically, but we cannot identify precisely.
Now, in terms of the variables that we can measure directly, we can only explain around seven percentage points. And so it's 13 percentage points of that variation. We know it's there, but we can't actually put our finger out on it.
In terms of the variables that we do know make a difference, that is, contribute to the 7%, the school sector is important. The second thing that's really important is the academic orientation of the school. So if it's a school where the parents really care about the academic progress, those schools tend to do better. We also find something that's quite interesting, and this was a little bit difficult to explain. That it does seem that some factors that make schools different play a role.
So for example, we did find that being a single sex school was to some advantage compared to co-educational. But the problem is that there are very few single sex schools, so it's a little bit hard to tell whether it is the fact that it's single sex, or whether the fact that it's just different and schools that are just a little bit different obviously attract a somewhat different clientele. Similarly, we found that things like the use of streaming made a difference. So, schools that had no streaming at all did a bit better. But they've had any of those almost all schools do some sort of streaming. So again, it's something special about the school rather than about the streaming.
Size and resources do have some impact. So bigger schools on average do a little bit better. I guess the reason behind that is probably to do that there are economies of scale. You can have a richer curriculum with a bigger school. The principal has more ways of using their resources, so they get some flexibility.
Steve Davis (04:56)
I also was intrigued to see that, and ask you actually, what insights you could draw from the fact that schools with a higher performance of students from non-English speaking backgrounds, it tended to have an increased probability of going on to university than those from other schools.
Tom Karmel (05:16)
This is consistent with a lot of research. Now, when you're talking about people from a non-English speaking background, you do have to be careful because you're talking about people from a very wide range of backgrounds. But we get this result that because there are certain large migrant groups where the parents just put a huge importance on the education of their children. So it's got nothing to do with the education of the parents or the occupation of the parents or how rich they are. It's just they put a lot of importance on the education, and you see that flow through to the performance of their children.
Steve Davis (05:51)
In the report, I noticed that the terminology of high performance schools, average performance schools and low performance schools was used as you were looking through the data sets. One thing I noticed is that in the low performance category, it was only public schools that sat in that category, as opposed to a mix in some of the others.
Is it fair to say that students being enrolled in public schools that sit in this low performance category are starting off at a considerable disadvantage to fellow students who are going to higher average performance schools.
Tom Karmel (06:28)
This is a very difficult question. The thing to remember when we look at these distributions, we are controlling for individual characteristics. So what we find once we abstract from these average effects I was talking about before is that we get a distribution of performance, whether it's in TER or probability of going to university having controlled for TER. And so we can rank schools from performing quite badly once we control for the individual characteristics to very well.
And what you do find is you find that in the government sector, in the Catholic sector and in the independent sector, you get a spread of schools from poor performance to the high performance. But what the data does say is that there is a concentration among the government schools, more so at the bottom end of that distribution. So, it is going to be tough for students going to these schools to get high TERs and go to university compared to their peers who might have gone to a school that's performing better.
Steve Davis (07:36)
Are we talking about quite a small range of difference from the highest to lowest performers? Or is it actually quite vast?
Tom Karmel (07:45)
We're talking about very significant differences between the high performers and the low performers. So for example, if we look at the raw scores we're looking at average TER scores of around about 90. This is an average score for a school, not for the best student at the school, down to an average score of around about 50. We are talking about huge differences in the performance of schools.
Steve Davis (08:16)
Do you get a sense that these things that are so hard to measure, which in the report you refer to as positive idiosyncratic factors which could be ethos, school, culture, etc.? Can you foresee us getting to a point to really be able to nail these, to identify them such that they could be replicated in lower performing schools to try and correct the record there?
Tom Karmel (08:40)
That's what you'd always like to do. I think one of the weaknesses of the data set that we're playing with here is that we don't have a measure of teacher quality. Everybody knows that teacher quality is really important and that good teachers get good results. But it is very difficult to put your finger on precisely what makes for good teachers.
I guess my take on it is that you need schools where everybody takes responsibility for what the school is doing, and this gets back to the importance of educational leadership. I think the only way that you can really replicate good behavior right across the board is to have really strong educational leaders in all schools.
Steve Davis (09:23)
Just thinking further about these idiosyncratic factors pertaining to certain schools. Just how significant are they on a child's path, a student's path to university?
Tom Karmel (09:35)
Well, I noted that they explained around 13 percentage points of the variation. So we've got seven percentage points explained by the things we can measure 13 percentage points by these idiosyncratic components. Translating that into TER scores, we're talking about 15 TER points. So that's the difference between a score of about 70 and a score of 85. So you can see they're very important.
Steve Davis (10:04)
Thanks for listening to this podcast produced by the National Centre for Vocational Education Research. If you would like to download the The impact of schools on young people's transition to university, visit our website www.ncver.edu.au