Does not equal what they put here:
"It is argued that ethnic minority pupils have greater ambition, aspiration
, and work harder in school. This is the main argument here – London has more of these pupils and so has a higher average GCSE score than the rest of the country"
Sorry but that is correlation not proven causation. It is however being applied as and is presented as causative.
You are trying to argue that since correlation does not equal causation, correlation never derives form causation. That is flawed logic, correlation is a natural outcome of causative factors which is why it is one of the most fundamental tools in science and statistics. Professor Burgess has examined the data statistically and presented the results along with a hypothesis relating the a possible causal factor (ambition and aspiration) to the observed data. This is presented as an argument and not proof, hence the phrase "It is argued". That is typical across all fields of science: correlation studies can indicate links between variables and potential causal factors can be presented. Further evidence is needed to prove those causal factors in future research. Perhaps Professor Burgess will be researching those factors in future work, go ask him if you want.
Socioeconomic status and childhood performance (plus language ability, adult health) is also only linked by correlation with various causal factors postulated but unproven. Strong correlations exist and sound reasons are
argued, but they aren't proven except in a few specific cases.
Yes, I wonder why that is? Is it because the Birmingham results would disprove that it is not the fact they are from an ethnic minority but something else?
The Birmingham results are completely irrelevant to the study of London, as are results from Bangkok, Mumbai, Addis Ababa or Tashkent Uzbekistan. The study only researched London, why is that so hard for you to grasp?
Yes, I am looking for sound valid science that has a purpose not something looking for correlating factors that are most likely inconsequential
Most likely inconsequential in your opinion. This science does have a purpose, to investigate if the average London performance is better than the English average and what the main reasons for that effect are. The first sentence in the abstract sums up the goals perfectly:
This paper contributes to understanding the ‘London Effect’ , focussing on the role of the ethnic composition.
No, this research is aims to provide a pro-uncontrolled immigration argument. It fails to address known causative factors and uses a potentially limited to correlating factor to draw its conclusions. There is a reason for that - the research is biased. It reeks of it
Where is your evidence that the research is biased and driven by an agenda? If you have evidence it is biased then you can let Bristol University know because I am sure they wont want to be associated with biased research or unsound science. It doesn't try to address causative factors so it can't fail to address them. Furthermore, your argument is things like socioeconomic status of parents have a significant effect that is not accounted for - that effect is only known from correlations, there are no proven causation either.
Here is an example:
People in India and China eat rice!
People in Ireland eat potatoes!
People in India and China predominately have brown eyes!
People in Ireland predominately have blue eyes!
Therefore, rice makes you eyes go brown and potatoes makes you eyes blue!
Daily Mail "Increase in Chinese eye problems linked to consumption of rice!"
Again, you are trying to postulate that correlation is never an outcome of causative factors. That is flawed reasoning, correlation naturally follows from causation (and other non-causative coincidental factors).
But the London effect is better explained by the factor we already know to create a difference and that will still hold to be true across other cities that show different patterns to London ...
Is the London effect better explained by other factors? Do you have proof of that? Furthermore, do you have proof that these other factor are statistically independent of the ethnic background?
Your opinion is that school performance correlates positively with parental socioeconomic status. That is fine, the research doesn't seek to disprove that, in fact that might be one of the mechanisms that accounts for the results. Socioeconomic status also correlates positively with ethnic background/immigration status in London. Therefore ethnic background correlates with child school performance in London, which is supported by the data. You seem hellbent to recurse to a deeper level of research to prove why children from different ethnic backgrounds perform differently. That is valid research but is irrelevant to the the work done at Bristol University.
The paper is flawed because it tells us nothing meaningful or useful and has been produced at a time when its conclusions could be used for political gain. The fact is phrases its limited and meaningless remit to include that is interesting.
Even if the results are not meaningful to you personally doesn't make them flawed. They are a high level analysis of demographic differences in London compared to the rest of England. Not least the paper has provided additional evidence to support the existence of the London effect.
The timing of the research is just a natural outcome due to the "London effect" receiving increased research recently. The research was made in response to reports from the CFBT and IFS,
http://cdn.cfbt.com/~/media/cfbtcorporate/files/research/2014/r-london-schools-2014.pdf
Although there is substantial evidence of improvement in London schools, there is much less evidence of the causes of impact
The research by Prof Burgess is just work along the same direction, reaffirming the phenomenon exists and this time finding important ethnic differences in average performance that help explain the results. To explain why the immigrant children do so well, yes, you will have to look at other
factors
