Word_Swordsman said:
The "adjustments" are adding a huge amount of subjective opinions that twist the data.
No they don't. The variables are pretty objective for the study and all would be important variables when considering education levels among schools.
There can't be anything wrong with averaging SAT or ACT scores between the school groups.
So how do you suggest controlling for heterogeneity?
There's nothing wrong with doing a study with SAT/ACT scores to look at different school types, btw. But you still have to use appropriate statistical techniques to analyze the data.
Do that first, report the actual raw data summary, then apply various adjustments to change the analysis to one's heart's desire, but leave the data alone.
The data in the study was "left alone." Raw data doesn't tell you much. But using raw data to prove a hypothesis without utilizing appropriate statistical methods would establish that you don't know statistics. Raw data does not establish causal connections or relationships among variables.
I don't believe family background changes a school's curriculum for a whole classroom. Public schools follow government curriculum, private schools being required only to meet minimum state requirements. Some public schools have stopped use of test score numbers, even forbidding the scoring of ball games.
Ok.
There is no way any sensible person can demonstrate, for instance, an inner city failing school in a broken district full of dropouts, plagued with violence against teachers, with relatively few students able to read or do simple math is statistically equal to an elite private school producing a very high percentage of graduates averaging very high SAT/ACT scores, successful employment, professionals, etc.
Has anyone argued to the contrary? In any case, so you just listed a bunch of variables that affect school performance: dropouts (ie, graduation rates), violent acts, and a "broken district" (whatever that is exactly). So if you were more interested in factors that affect performance rather than simply theorizing about them, you could test those variables empirically to estimate a size of effect. Previously on this thread, it was suggested that a decline in education levels was caused by the removal or prayer and bibles from schools. This is empiraclly testable and is a totally separate question from whether or not private schools outperform public schools.
Yes, private schools DO outperform public schools. The study I linked to on this thread even says as much, so I'm not sure why you keep bringing it up. The interesting question then is to figure out why. Is it because public schools banned prayer/bibles? I doubt it. If you want to argue in support of that though, by all means, do so.
Of course, manipulation of data through enough statistical analysis techniques can be made to demonstrate 'whatever', "adjusted" by parameters that don't fit all schools sampled. The parameters listed in that study indeed compares apples to prunes to bananas, not school type to school type. All the school types are obviously made the same type through manipulation of the data. Nobody with a lick of sense believes they are all the same. The approach of the study was skewed, even if they used typical statistical tools.
It's not "skewed" or "manipulated." By repeating this assertion, it's clear to me that it's because you probably just don't know how to do statistical analysis.
You can use statistical techniques to prove that someone's data is biased in some way though. But you haven't done this.
Our public schools are sold on Common Core, leaving parents unable to help their kids with homework. They can't find out how the teachers are trying to teach kids to do math. We haven't found one student that can do simple math without a bar graph on paper, or rows and columns to add, subtract, etc. But we know the private/church schools are teaching the old tried and true methods of rational thinking, and what the state requires. That's just one example of blindly following D of E guidelines, or doing what has been working.
The differences between the schools are huge! But our government is blind to that.
I dunno, this sounds like repetition of partisan politics to me. I'll leave this to someone else since I was only concerned with one claim on this thread.
