Education, politics, and anything else that catches my attention.
Sunday, March 26, 2006
Good Students vs. Bad Students
I was reading a post on RightWingProf's blog and came across this statement, with which I cannot help but agree:
Oh yes, there are bad students. There are bad teachers, too. But we all had bad teachers. The difference between a good student and a bad student is that a good student does not use his teacher as an excuse for being too lazy to do the work.
Actually, that was what I was reading when I clicked over to see if I had any comments to publish on my own blog :-) Ah, the things we math geeks do on a Sunday morning.
To be honest, I had difficulty following your analysis because I didn't understand enough about the raw data that you analyzed. What data did you use, and what exactly do you assert your analysis of it means?
I posted the excel file that contains the data. There are three variables for each state: per pupil spending, %age of students at or above proficiency in math, and %age of students at or above proficiency in reading. I first wanted to see if there was a statistically significant relationship between the two proficiency scores, so I ran Anova on the data and found no statistically significant relationship. That's not too surprising when you think about it, actually. One state could have 30% at or above proficiency in reading, but 10% in math and vice versa.
Then I did linear regressions to see if you could predict the reading and math proficiencies from the per pupil spending. Nope. There was hardly even a correlation (without looking, I think the R value was 0.06 on one and 0.04 on the other). Certainly not good for those who insist that the only way to teach students and get them to pass standardized exams is to shell out more money.
I could have run non-linear regressions but I didn't (maybe I'll do that when I get some free time).
ANOVA: Analysis of variance (see here), a statistical test used to determine if the variance between two or more means is or is not statistically significant.
And the quality of life analysis is posted on my blog now.
blush
ReplyDeleteYou might find this interesting, math and school spending.
Actually, that was what I was reading when I clicked over to see if I had any comments to publish on my own blog :-) Ah, the things we math geeks do on a Sunday morning.
ReplyDeleteTo be honest, I had difficulty following your analysis because I didn't understand enough about the raw data that you analyzed. What data did you use, and what exactly do you assert your analysis of it means?
I posted the excel file that contains the data. There are three variables for each state: per pupil spending, %age of students at or above proficiency in math, and %age of students at or above proficiency in reading. I first wanted to see if there was a statistically significant relationship between the two proficiency scores, so I ran Anova on the data and found no statistically significant relationship. That's not too surprising when you think about it, actually. One state could have 30% at or above proficiency in reading, but 10% in math and vice versa.
ReplyDeleteThen I did linear regressions to see if you could predict the reading and math proficiencies from the per pupil spending. Nope. There was hardly even a correlation (without looking, I think the R value was 0.06 on one and 0.04 on the other). Certainly not good for those who insist that the only way to teach students and get them to pass standardized exams is to shell out more money.
I could have run non-linear regressions but I didn't (maybe I'll do that when I get some free time).
What exactly is Anova?
ReplyDeleteAnd did you account for the wildly different costs of living in the states (South Dakota vs. Connecticut or California), for example?
No. It wasn't in the data. That would be interesting, though, as would be teacher student ratio, etc.
ReplyDeleteANOVA: Analysis of variance (see here), a statistical test used to determine if the variance between two or more means is or is not statistically significant.
ReplyDeleteAnd the quality of life analysis is posted on my blog now.