Comments on: Test analyst: Reading exam bar even lower than critics say

By: C Sitsen

C Sitsen — Mon, 24 Aug 2009 13:28:18 +0000

Thanks for an excellent article. Does the State Education Department have any thoughts on this? Additionally, the State might want to tell us how and when the standards were set.
Mr. Smith’s continual urging fior an ombudsman must be taken seriously. Parents and students need and deserve an advocate.

By: Fred Wildman

Fred Wildman — Sat, 22 Aug 2009 15:46:01 +0000

Thank you for providing a cogent and statistically accurate summary of some of Fred Smith’s insights into the use of these statewide tests for a purpose for which they were apparently never intended. I am especially delighted that you corroborated Smith’s analysis with the results of Professor Corter’s calculations of the probability of achieving a particular raw score without a student having any knowledge and simply guessing. For those who doubt either of these two experts I suggest utilizing a table of probabilities associated with the binomial distribution (i.e., the mathematical distribution which deals with two and only two outcomes; in this case success or failure on a given multiple choice item) and all disbelief shall be washed away. These somewhat subtle mathematical realities are easily overlooked by those for whom the political rather than the educational merits are the coin of the realm. In the nature of disclosure I was a colleague of Mr. Smith during the the early 1980s when we bought were employees of the (former) NYC Board of Education.

By: Robert Pondiscio

Robert Pondiscio — Sat, 22 Aug 2009 03:40:13 +0000

Do you have a link to Smith’s results?

By: Maura Walz

Maura Walz — Sat, 22 Aug 2009 01:21:25 +0000

Fred Smith looked at the ELA exams in grades 5, 6 and 7 and reported the same pattern in those tests. Prof. Corter’s calculations apply only to the fifth grade ELA exam. The calculations are based on the specific number of multiple choice questions on a test, so it will vary from exam to exam.

By: Leonie Haimson

Leonie Haimson — Sat, 22 Aug 2009 00:04:21 +0000

Question: are Fred Smith’s findings true of all the state exams 3-8 in all subjects; or just some? and what about the probability findings of Prof. Corter?

By: Diana Senechal

Diana Senechal — Fri, 21 Aug 2009 23:27:49 +0000

I have been doing some more experiments since the initial one, and my results coincide with Smith’s. I am delighted that Smith has been looking at the scoring of the written portions. It is indeed easy to score a few points on those, and with 2 or 3 points in hand, the chances of passing by guessing are very high indeed. Also, I found that the number and proportion of students at level 1 is lowest where the odds of attaining a 2 by guessing are highest (i.e., on the third grade math exam, the fifth, sixth, and seventh grade ELA exams, and the seventh grade math exam).