DfE finds that higher parental incomes buy better educational outcomes

[noblegiraffe]

In a piece of research that will surprise no one, it turns out that children of wealthier parents do better at school.

However, while it is obvious that PP students and especially FSM pupils perform particularly badly, pupils from below-median-income families perform lower than, but more in line with children from wealthier families than with PP pupils.

What the DfE really want to know in this consultation, however, is whether they should refer to below-median-income families who don't qualify for PP as 'Ordinary Working Families'.

consult.education.gov.uk/school-leadership-analysis-unit/analysing-family-circumstances-and-education-1/

Good to know that they are spending their time and effort focusing on the key issues in education at the moment.

clarity I think noble has been more than patient with you on this thread. Please don't exhaust her too much!

PP is to close the gap for a broad group of students.
Lots of little interventions can go towards larger successes.
Data is starting to show which schools are managing to close the gap.
They can share good practice with others.
Over time other schools will report what seems to work for them.
It is a 5 year process at secondary. The interventions done in y7 may not be shown to bare fruit until y11.

What you are asking for is impractical.

Did my DD1 pass English language because of intervention in y9, or intervention in y11, or because of the revision sessions, or support from home? Which of these raised her marks by 4, or by only 3 or by 6? Ultimately you have no way of knowing. You only know that as a whole they were successful.

It doesn't take an expert to pick holes in what you are saying.

You need to listen to Dylan Wiliam. Here are his slides from a talk he gave called 'Why teaching will never be a research-based profession'

www.dylanwiliam.org/Dylan_Wiliams_website/Presentations_files/2016-10-29%20ResearchED.pptx

clarity I think noble has been more than patient with you on this thread. Please don't exhaust her too much!

Who is exhausting who? noble is under no obligation to respond to my comments or answer my questions, Tween.

Over time other schools will report what seems to work for them.

How will they determine this, Tween?

What you are asking for is impractical.

Why? The data I mentioned is recorded anyway. Schools could undertake a trial analysis of it, to see if they find this helpful and whether progress and attainment in the PP group improved.

You need to listen to Dylan Wiliam. Here are his slides from a talk he gave called 'Why teaching will never be a research-based profession

So why do educational professionals talk so highly about their 'evidence based approaches'?

Because some educational professionals don't understand data, some make up their research and some actually run proper trials with some attempt at controlling for other variables.

Evidently, noble...

?

Well, not understanding data and making up research is not exactly going to help bridge gaps in attainment between different sectors of pupils, is it, noble? So I was correct in saying it was worth considering whether the actual targeting of the PP group for interventions actually put them at a disadvantage?

This is the sort of approach is more reasonable than a guy at a school with a spreadsheet:

"One of the most promising projects we’ve funded was an initiative delivered by the Calderdale Effectiveness Partnership that cost just over £50 per pupil. Designed to use self- regulation to improve writing skills, the project provided children with memorable experiences such as a trip to zoo, and gave them a structured approach to writing about it. Pupils made, on average, an additional nine months' progress; the impact on free school meals pupils was even greater, at 18 months.
To assess its impact as rigorously as possible, the evaluation was set up as a randomised controlled trial led by an independent evaluation team. We’re now testing the project’s effectiveness on a larger scale, working with 7,200 pupils in Leeds and Lincolnshire, and are hugely excited by its potential."

www.suttontrust.com/wp-content/uploads/2015/06/Pupil-Premium-Summit-Report-FINAL-EDIT.pdf

In the report it shows that the attainment 8 gap for disadvantaged pupils appears to be closing, which would be hard to explain if PP money was actually making pupils worse.

The project you mention, noble, according to the information you have given, notes clearly which pupils were included and the nature of the intervention they received. It then, I presume, analyses progress in each of the separate groups and compares outcomes to starting points and predicted progress trajectories, in order to extrapolate additional progress made.

The same type of study, although not quite as scientifically with regard to randomised controlled trials, could be done from data schools already hold. Surely this would be better than schools merely reporting 'what seems to work for them'?

No, you have to set out in advance what you are looking to find, then run your trial and then analyse your data.

Fishing about in data that you have already collected in order to look for positive correlations as yet unknown is statistically dodgy. If you look long and hard enough at enough subgroups you will find interventions which give better than average results, and interventions which give worse results.

Scientists do occasionally do this sort of population analysis, but they have to be very careful about it and use much larger samples than the average school would yield.

No, you have to set out in advance what you are looking to find, then run your trial and then analyse your data.

This could be done. I was not talking about necessarily analysing data retrospectively, I was just referring to the fact schools are well practiced in collecting the type of data that could be utilised in this type of study. This data need not necessarily be analysed exclusively by individual schools (although they could do their own informal analysis), either. Schools could submit their data for analysis by a centralised body, in a way similar to what happens with the school's census.

You know that people do this sort of thing for a living, right? That well-meaning amateurs would just make a hash of things?

What you actually want is for well thought-out RCTs to be conducted by people who know what they are doing. Which is not schools.

What you actually want is the sort of thing I listed above. Which are already going on.

noble, so you are telling me, schools, when tasked with narrowing the attainment gap between PP qualifying students and those who don't qualify for PP, within their own school, should not even attempt to systematically evaluate the success of interventions they have funded through PP?

That they should not be proactive in seeking to use the data, they already are used to collecting, in order to inform what interventions might be the most effective and instead should effectively decide on which interventions to start or continue from a vague feeling of 'what works for them'?

Furthermore there is no need to be too conscientious regarding systematically identifying the needs of the PP qualifying group or ensuring the PP funding is targeted towards meeting their actual needs? Since the evaluation process will take longer than a large proportion of these students' school careers?

No wonder progress is slow...

The Sutton Trust report I linked to upthread says that schools are not expected to start with a blank slate. Obviously schools and teachers have experience in improving pupil outcomes and can draw on this. There's also the teacher toolkit summary of educational evidence: educationendowmentfoundation.org.uk/resources/teaching-learning-toolkit

But you're right, schools on their own won't be able to systematically evaluate the impact of any interventions in an especially meaningful way. They can survey pupils (did you find this useful?) they can look at exam data (bearing in mind recent research showing that results are volatile simply because each year has a different cohort of students cambridgeassessment.org.uk/news/fluctuations-in-schools-exam-results-normal/ ) but they shouldn't expect to be able to reliably say 'this intervention was more cost effective than this one'.

And even when the results of RCTs come out, remember Dylan William from above - if you reproduce the study, you'll probably get different results. This problem has hit Carol Dweck's Growth Mindset theory which initially looked promising: people can't replicate her results.

In the report it shows that the attainment 8 gap for disadvantaged pupils appears to be closing, which would be hard to explain if PP money was actually making pupils worse.
Unfortunately, though, from what I recall, attainment 8 was only introduced relatively recently - so the goal posts have been moved haven't they?

But you're right, schools on their own won't be able to systematically evaluate the impact of any interventions in an especially meaningful way.
Floundering in the dark and clutching a straws are phrases that come to mind. So is it really better to attempt to target the PP qualifying group with such interventions, instead of concentrating on identifying any individual student's additional educational needs, tailoring teaching accordingly and analysing progress?

Attainment 8 is GCSE point score in 8 subjects in various categories. It doesn't matter that the goal posts have changed because they changed for all pupils at the same time and what's being looked at is the gap, which is narrowing, not overall score.

How would we tailor teaching according to an individual's additional needs without extra money to do so? And the flag to highlight that they probably need extra support?

How would we tailor teaching according to an individual's additional needs without extra money to do so? And the flag to highlight that they probably need extra support?

In the same way as additional needs were tackled before PP came in.

Every school should be putting money aside for low level additional needs, as part of the SEN budget. It is only when 6k has been spent on an individual child that High Needs Funding is triggered and an assessment for an ECHP takes place.

In addition to this, to account for needs that are clearly more economic, such as not being able to afford school trips, schools which are in socioeconomically deprived areas receive a higher level of AWPU funding. The amount of families claiming FSM is usually one of the factors which determines this.

If there's one thing we've learned from the grammar school debate it's that schools in socioeconomically deprived areas do not necessarily have a correspondingly deprived intake.

Identifying those disadvantaged pupils at-risk of educational underachievement and allocating them the money is far fairer, and enables schools to be proactive rather than reactive.

If there's one thing we've learned from the grammar school debate it's that schools in socioeconomically deprived areas do not necessarily have a correspondingly deprived intake.

Well, as discussed in my earlier post PP funding can create and perpetuate the same socioeconomic divides grammar schools do, by creating pseudo-grammar schools through different means. As I earlier explained, this can happen when PP qualifying factors are conflated with SEN and consequently parents, whose children do not qualify for PP funding, move their children to schools which receive a lot of PP funding, simply because the PP is used to fund interventions which are commonly used to tackle SEN.

Identifying those disadvantaged pupils at-risk of educational underachievement and allocating them the money is far fairer, and enables schools to be proactive rather than reactive.

Is it? Or is it just a mechanism for promoting prejudice and low aspirations? I think you are talking about more than being proactive, you are talking about trying to predict the future. Not that I believe in being reactive, either. Being reactive suggests someone is taking an action, without really thinking it through. I would prefer educational professionals to be responsive, that is, take an appropriate, reasoned action in response to a given occurrence.

So in the situation outlined above, you get:

School A: a high level of PP funding
a high occurrence of interventions commonly used to tackle SEN funded through PP.

children with SEN, without an ECHP who do not receive PP, moving to this school
= a large proportion of children at the school face barriers to their achievement.

School B: a low level of PP funding
a low occurrence of interventions commonly used to tackle SEN funded through PP
children with SEN, without an ECHP, who do not receive PP, less likely to move to the school
= a small proportion of children at the school face barriers to their achievement.

You're confusing anecdote with data again, and correlation with causation.

You're saying this (situation outlined above) does not happen, noble? I've read widely of this happening and I've personal experience of knowing parents who have moved their child with SEN for this reason. It was mooted, as a possibility, to me, when my DC was going through the Statementing process by the SENCO and Deputy head of his school.

What I would like is truly comprehensive education. Comprehensive schools which are geared towards properly serving the needs of the local community and places which would be allocated to local children. No child should have to travel great distances to school because of barriers to their learning or the lack of them. And quite frankly people should be happy to have their children educated alongside the children of the community they live amongst. This would not cloud the issue of which schools might require a higher AWPU either.

You're confusing anecdote with data again, and correlation with causation.

And there's no chance of schools doing this, if they refuse to consider even informally analysing, the data they routinely collect and instead rely on vague feelings 'of what works'? It is a very convenient excuse not to do anything, about anything, if you have to wait years for an 'expert' to tell you how to act. After all, science is incomplete, there is no Universal Theory, should we ignore all the conclusions that we, so far, have been able to make?

This is interesting, it is a filled in template for a Pupil Premium Review:

https://www.tscouncil.org.uk/wp-content/uploads/2016/12/PP-strategy-template-secondary-example-FINAL-1.docx

Some of the analysis,I was talking about is mentioned here, for example,

"...^meet the success criteria? Include impact on pupils not eligible for PP, if appropriate.
Lessons learned^
(and whether you will continue with this approach)
Cost
Improved Year 8 literacy results
One to one tuition delivered by qualified teacher
High: observed increased progress amongst participating children compared to peers, as measured using scores on the Progress in English test.
Success criteria: met."

This actually talks about "including impact on pupils not eligible for PP," and comparing participating children to peers.

So, I was not talking complete rubbish. Analysis of the data, in the way I outlined is possible and others must be finding it useful for the above link to be suggested from the government website. It also says that reviews are undertaken either because OFSTED flags up a problem or that schools decide to do this as part of their own evaluation process.