DC-CAS
![[Post New]](/jforum/templates/default/images/icon_minipost_new.gif){kind=icon}
Anonymous
| Does any know what a child has to score to be proficient? |
Anonymous
My understanding - but someone please correct me! - is that it's measured in quartiles based on the actual distribution of student scores (of all DC public and charter school students at that grade level). By asking what score is deemed proficient, you're actually asking what score does the median student have this year. The 50% of students above that score are deemed "proficient", the top 25% of which "advanced". The 50% of students below that score are deemed "basics" or "below basics". Ever seen a boxplot? That's it. So the cutoffs are empirical figures that change with every test. You can easily derive that they get harder to reach when all students get better or better students arrive in the mix. And just so you don't get your hopes up, the kind of question your asking here, about distributions, likelihoods and such, or versions thereof can well be found on 4th or 5th grade DC-CAS tests... |
Anonymous
| Really? There's not a set score for proficient etc? |
Anonymous
| 10:00 is wrong. There is a set score for proficient. |
Anonymous
10:00 here: Glad to stand corrected but what would that "set" score be, please? Or maybe we actually mean the same thing? What I'm saying is that, say, the 5th grade DC-CAS test comprises a number of questions yielding a maximum of 100 points, and 25% of the test takers score at least 68 out of 100, 50% score at least 54, and 75% score 36, then the "set" score for "advanced" in 5th grade that year is 68. A score of 54 to 67 gets you a "proficient" rating. And a score of 36 to 53 is rated basics. Anything below 36 is below basics. Maybe what I'm getting wrong is that it's not as empirical as that and the "set" scores are predefined on the basis of expected or least year's quartiles rather than actually observed quartiles. But unless someone gives me better details then I'm quite certain that this is more or less how it works. Of course, my statistical conscience would much prefer averages than quartiles. In the above example, scoring 54 or scoring 67 is actually a huge difference. And schools that have a large portion of students below proficient may have a lot just below the 54 mark or they may have a lot near the 36 mark, which is a whole different story. |
Anonymous
|
The point of NCLB is to get all students above the bar to proficient. If the quartiles were defined after the test, it would be impossible, because the bottom quartile would always be below basic, etc.
I believe that the test makers have defined the passing grade / cut off between below basic / basic/ proficient / advanced, when the test is written. Tests are scored and students fall into the bins they fall into. |
Anonymous
| 19:22 back. I've only actually seen my child's 3rd grade DC-CAS, where the break between prof / advanced was somewhere around 80. But without seeing the test and how it is scored, I have no way of interpreting what an "80" means. The tests are not publicly released, just a sampling of questions. |
Anonymous
| My guess is that DCPS decides what is proficient when they get the data. They are probably doing that now, figuring out how to massage the data so that they have some good news to report. |
Anonymous
|
When I received my child's DC CAS scores last August, I calculated that 67% in reading and 68% in math was proficient. The year before it was 68% for reading and 71% for math (5th grade).
So it could be that the box system with quartiles is being used. |
Anonymous
That doesn't make any sense at all. If DCPS could set the proficiency marker, they wouldn't be condoning rampant cheating to raise scores. |
Anonymous
I don't follow your logic, but then, when it comes to DCPS, or education policy in general, logic has nothing to do with it. |
Anonymous
If DCPS set the proficiency rate, they would set it much lower to improve their reputation. I don't understand how anyone would think DCPS has this power, it wouldn't be logical. |
Anonymous
I think that OSSE is suppose to set the benchmarks. But OSSE hasn't shared what those benchmarks are with school staff. |
Anonymous
|
Isn't OSSE and DCPS of the same cloth. If you don't think there's NOT any serious discussion and surrendering regarding test scores. THEN YOUR LOGIC is loco! Remember all of these employees are appointed and hence there's a loyalty factor to the politicians and not the pupils.
As for the rampant cheating if the bar is set low and the gains are so easily attainable, well it doesn't take rocket science to figure it out. A bad conscientous and cash is all one need as incentive too cheat. |
Anonymous
No, OSSE and DCPS are totally separate. OSSE is the Office of the State Superintendent of Education, the office where NCLB and RTTT grants are administered. OSSE determines what is on the DC CAS and it's up to OSSE to determine what the benchmarks are for advanced, proficient, basic, and below basic for all tested grade levels. |