Thursday, May 15, 2008
Wednesday, May 14, 2008
What Does it Mean?
The outcome measure is student score on a standardized test. A norm-referenced test. That means that the test scores can be compared from year to year to get a sense of how much a student has learned. There is no "ceiling" on the score, so there is always growth.
In my case, I am really interested in the month-by-month rate of growth in achievement, and how much specific kinds of teaching might influence that growth. So these equations include terms to test different kinds of instruction.
In addition, there are terms that account for "balance" between kids who get the instruction and kids who don't. That way this mimics a true experiment. Think of medicine, where some people get the real drug and some get a placebo. The study is randomized, or people are matched on characteristics, so researchers can look for how the drug works by comparing to people without the drug. That's what I did statistically....create a kind of "matched" set of students who got it and didn't.
My head hurts just thinking about it. Six days to go.
In my case, I am really interested in the month-by-month rate of growth in achievement, and how much specific kinds of teaching might influence that growth. So these equations include terms to test different kinds of instruction.
In addition, there are terms that account for "balance" between kids who get the instruction and kids who don't. That way this mimics a true experiment. Think of medicine, where some people get the real drug and some get a placebo. The study is randomized, or people are matched on characteristics, so researchers can look for how the drug works by comparing to people without the drug. That's what I did statistically....create a kind of "matched" set of students who got it and didn't.
My head hurts just thinking about it. Six days to go.
Sunday, May 11, 2008
My Equations
I'm updating my statistical equations today, in my last revisions on the thesis. Just for fun, I thought I'd share them. Looking at them now, I get kind of queasy. And I know what they mean.
[6] Υtij = π0ij + πij(month) + π2ij(month2) + etij
[7] π0ij = β00j + β01j(male student) + β02j(minority student) + β03j(chronological age) + β04j(student SES) + rij
π 1ij = β10j + β11j(male student) + β12j(minority student) + β13j(chronological age) + β14j(student SES)
π2ij = β20j + β21j(male student) + β22j(minority student)
[8] β00j = γ000 + γ001(higher writing/higher reading) + γ002(higher writing/medium reading) + γ003(higher writing/lower reading) + γ004(medium writing/higher reading) + γ005(medium writing/medium reading) + γ006(medium writing/lower reading) + γ007(lower writing/higher reading) + γ008(lower writing/medium reading) + γ009(reading strata 1) + γ010(reading strata 2) + γ011(reading strata 3) + γ012(reading strata 4) + γ013(reading strata 5) + γ014(reading strata 6) + γ015(reading strata 7) + γ016(reading strata 8) + γ017(reading strata 9) + γ018(reading strata 10) + γ019(reading strata 11) + γ020(writing strata 1) + γ021(writing strata 2) + γ022(writing strata 3) + γ023(writing strata 4) + γ024(writing strata 5) + γ025(writing strata 6) + γ026(writing strata 7) + γ027(writing strata 8) + γ028(writing strata 9) + γ029(writing strata 10) + γ030(writing strata 11) + u00
β01j = γ010
β02j = γ020
β03j = γ030
β04j = γ040
β10j = γ100 + γ101(higher writing/higher reading) + γ102(higher writing/medium reading) + γ103(higher writing/lower reading) + γ104(medium writing/higher reading) + γ105(medium writing/medium reading) + γ106(medium writing/lower reading) + γ107(lower writing/higher reading) + γ108(lower writing/medium reading) + γ109(reading strata 1) + γ1010(reading strata 2) + γ1011(reading strata 3) + γ1012(reading strata 4) + γ1013(reading strata 5) + γ1014(reading strata 6) + γ1015(reading strata 7) + γ1016(reading strata 8) + γ1017(reading strata 9) + γ1018(reading strata 10) + γ1019(reading strata 11) + γ1020(writing strata 1) + γ1021(writing strata 2) + γ1022(writing strata 3) + γ1023(writing strata 4) + γ1024(writing strata 5) + γ1025(writing strata 6) + γ1026(writing strata 7) + γ1027(writing strata 8) + γ1028(writing strata 9) + γ1029(writing strata 10) + γ1030(writing strata 11) + u00
β11j = γ110
β12j = γ120
β13j = γ130
β14j = γ140
β20j = γ200
β21j = γ210β22j = γ220
[6] Υtij = π0ij + πij(month) + π2ij(month2) + etij
[7] π0ij = β00j + β01j(male student) + β02j(minority student) + β03j(chronological age) + β04j(student SES) + rij
π 1ij = β10j + β11j(male student) + β12j(minority student) + β13j(chronological age) + β14j(student SES)
π2ij = β20j + β21j(male student) + β22j(minority student)
[8] β00j = γ000 + γ001(higher writing/higher reading) + γ002(higher writing/medium reading) + γ003(higher writing/lower reading) + γ004(medium writing/higher reading) + γ005(medium writing/medium reading) + γ006(medium writing/lower reading) + γ007(lower writing/higher reading) + γ008(lower writing/medium reading) + γ009(reading strata 1) + γ010(reading strata 2) + γ011(reading strata 3) + γ012(reading strata 4) + γ013(reading strata 5) + γ014(reading strata 6) + γ015(reading strata 7) + γ016(reading strata 8) + γ017(reading strata 9) + γ018(reading strata 10) + γ019(reading strata 11) + γ020(writing strata 1) + γ021(writing strata 2) + γ022(writing strata 3) + γ023(writing strata 4) + γ024(writing strata 5) + γ025(writing strata 6) + γ026(writing strata 7) + γ027(writing strata 8) + γ028(writing strata 9) + γ029(writing strata 10) + γ030(writing strata 11) + u00
β01j = γ010
β02j = γ020
β03j = γ030
β04j = γ040
β10j = γ100 + γ101(higher writing/higher reading) + γ102(higher writing/medium reading) + γ103(higher writing/lower reading) + γ104(medium writing/higher reading) + γ105(medium writing/medium reading) + γ106(medium writing/lower reading) + γ107(lower writing/higher reading) + γ108(lower writing/medium reading) + γ109(reading strata 1) + γ1010(reading strata 2) + γ1011(reading strata 3) + γ1012(reading strata 4) + γ1013(reading strata 5) + γ1014(reading strata 6) + γ1015(reading strata 7) + γ1016(reading strata 8) + γ1017(reading strata 9) + γ1018(reading strata 10) + γ1019(reading strata 11) + γ1020(writing strata 1) + γ1021(writing strata 2) + γ1022(writing strata 3) + γ1023(writing strata 4) + γ1024(writing strata 5) + γ1025(writing strata 6) + γ1026(writing strata 7) + γ1027(writing strata 8) + γ1028(writing strata 9) + γ1029(writing strata 10) + γ1030(writing strata 11) + u00
β11j = γ110
β12j = γ120
β13j = γ130
β14j = γ140
β20j = γ200
β21j = γ210β22j = γ220
Tuesday, May 06, 2008
More Dissertation
It isn't over. Maybe it's never over. The joy of graduation, the giddiness and silliness that follows is now replaced by more slogging through revisions. Sigh.
I have to get the final, final, final document to the powers-that-be on May 21. Maybe then it will be over.
I have to get the final, final, final document to the powers-that-be on May 21. Maybe then it will be over.
Tuesday, April 29, 2008
Monday, April 28, 2008
The End of Being a Student
Our graduation speaker Michelle Rhee, chancellor of the District of Columbia schools is one smart and controversial woman. Fortunately, she got her speech down to a couple of bullet points so the brain dead doctoral students could follow. (We are seriously burned out.)
She found when she took over the DC schools no accountability, and a culture centered around serving the adults in the system, not the kids. And she went after both, upsetting the status quo rather decisively, it would seem.
So, she was pretty strident in her concerns about student achievement, teacher quality, about how adults in the system fail children. She was also pretty blunt about not accepting the excuse that kids are poor, or minority, or otherwise challenged as the explanation for their academic failure. She would say, too bad they didn't have a great breakfast, you still have to teach them math. She challenges the disheartened to get out of the business to make room for the energized.
Surprisingly, she was interrupted by applause from the audience at least 6-7 times. And she was saying some pretty tough stuff. Student scores matter. Adult interests in job situations don't matter. And so on. I would bet that UM is the only Ed School that asked her to speakat graduation. But credit my dean, Deborah Lowenberg Ball. My dean is a very unusual woman. I'll have the privilege of working with her and others at the Ed School school for at least another year as a postdoc.
UMich school of education is really a special place, at least I think so. I came here seven years ago expecting to see another corner of the Edusphere dedicated to maintaining the status quo. What I found here is a vibrant institution where academics push to improve teaching and learning. Our work is slow and painstaking, and doesn't yield the instant solutions I wish it did. But doing research with care and caution, testing and retesting, leads to sound findings rather than fads and trends.
I'll be honest, when I came to Ed School I didn't expect to hear this at a graduation. I was quite enthused by her talk. So were my daughters. So were most in the audience. I am told by sources that the faculty felt challenged and inspired by her talk. But some doctoral students, or graduates by then, complained about her talk as they stood outside the hall after the ceremony. There was some grousing that she was too demanding, or that she failed to understand the nuances of student failure or something. I couldn't eavesdrop well enough to know.
I may post more photos from graduation.
I'm so tired I can't explain.
But I did it.
She found when she took over the DC schools no accountability, and a culture centered around serving the adults in the system, not the kids. And she went after both, upsetting the status quo rather decisively, it would seem.
So, she was pretty strident in her concerns about student achievement, teacher quality, about how adults in the system fail children. She was also pretty blunt about not accepting the excuse that kids are poor, or minority, or otherwise challenged as the explanation for their academic failure. She would say, too bad they didn't have a great breakfast, you still have to teach them math. She challenges the disheartened to get out of the business to make room for the energized.
Surprisingly, she was interrupted by applause from the audience at least 6-7 times. And she was saying some pretty tough stuff. Student scores matter. Adult interests in job situations don't matter. And so on. I would bet that UM is the only Ed School that asked her to speakat graduation. But credit my dean, Deborah Lowenberg Ball. My dean is a very unusual woman. I'll have the privilege of working with her and others at the Ed School school for at least another year as a postdoc.
UMich school of education is really a special place, at least I think so. I came here seven years ago expecting to see another corner of the Edusphere dedicated to maintaining the status quo. What I found here is a vibrant institution where academics push to improve teaching and learning. Our work is slow and painstaking, and doesn't yield the instant solutions I wish it did. But doing research with care and caution, testing and retesting, leads to sound findings rather than fads and trends.
I'll be honest, when I came to Ed School I didn't expect to hear this at a graduation. I was quite enthused by her talk. So were my daughters. So were most in the audience. I am told by sources that the faculty felt challenged and inspired by her talk. But some doctoral students, or graduates by then, complained about her talk as they stood outside the hall after the ceremony. There was some grousing that she was too demanding, or that she failed to understand the nuances of student failure or something. I couldn't eavesdrop well enough to know.
I may post more photos from graduation.
I'm so tired I can't explain.
But I did it.


