Reply to I just scooped the DCPCSB

Forum Index » DC Public and Public Charter Schools

Reply to "I just scooped the DCPCSB - 2018 tiers"

Subject:

Emoticons





More smilies

Text Color: Font: Close Marks

[quote=Anonymous][quote=Anonymous][quote=Anonymous]

Okay, I looked a bit more and now see what you are talking about.  The fact that the different groups have different growth percentiles doesn't say anything about those groups per se.  But it does show how white students, who in DC are almost entirely middle/high income, grow in school performance at a faster rate than the average black student in DC, who, on average, are significantly lower income.  This is the achievement gap, playing out over time as the two groups grow further apart.  So focusing on schools which are better able to actually show growth among lower income students, is an appropriate focus for the PMF.  I do wonder if they do an "in-group" growth percentile analysis?  This has to be in education statistics literature somewhere.

It's a very interesting point, thanks for noticing that![/quote]

I think it is a worthy goal to close the achievement gap by growing low income, black, special ed, etc. kids faster. The PMF is the basis for a lot of high stakes decisions -- closure, replication, etc. A Tier 2 school that is serving at-risk, low-income black kids in Ward 8 isn't going to be eligible for replication even if they are outperforming other schools in their area and their students are outperforming similar students. [/quote]

Agreed.  But if you look at the PCSB manual for how they calculate the MGP for grades 4-8, it does use growth compared to students who scored similarly on last years test:

"Progress Measure for grades 4-8
a. Measure: MGP captures the median growth of all public school students’
progress by comparing changes in students’ PARCC scores to changes made
by other students with similar score histories on the PARCC in the previous
year. Calculating MGP is a three-step process:
(1) A student growth percentile (SGP) is calculated for each student,
which shows how that student performed in this year’s assessment
compared with other DC students who had similar performance in the
previous year’s assessment. For example, if 20 students had a score of
340 in last year’s PARCC test, a student who did better than 15 of
those students in this year’s test would have an SGP of 75, since that
student did better than 75% of the students with a similar score on last
year’s assessment. Scores from all District students, including those at
DCPS schools, are used to determine an academic peer group and to
calculate SGPs.
(2) All of the students’ SGP scores for a school are arrayed from high to
low and the midpoint, or median, of these scores becomes the school’s
median growth percentile, or MGP. The higher the score, the more
students are improving compared with students attending other public
schools in the District.
(3) DC PCSB calculates a two-year weighted average (by n-size) by
averaging the school’s MGP values from two consecutive years. The
two-year weighted average is used to mitigate fluctuations in scores
from year to year. "

So this is an "apples to apples" comparison.  If my kid grew by 10 points, and kids who did similarly last year grew 10 points, my kid gets an SGP of 50.  But if my kid grew by 7, and kids who did similarly last year grew by 5 points, my kid gets an SGP of maybe 75 (? not sure how the math is done exactly), because he grew by more than similar students grew.

So, yes, white (i.e. richer on average in DC) students grow faster on average.  But since it's about each individual student's growth compared to the growth of similarly scoring students, that is taken into account.

I am not sure I am explaining it 100% correctly, let me know if that makes sense.

[/quote]

Options

	Disable HTML in this message
	Disable BB Code in this message
	Disable smilies in this message

Review message

Search Recent Topics Hottest Topics