<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="Content-Type"
content="text/html;charset=windows-1252">
<title></title>
</head>
<body>
Well, the only study is by me and is so far unpublished. I am not sure the
data I have would meet publication standards. It is always found to be interesting,
and a bit shocking, but I am not sure it predicts anything or explains anything,
except it clearly demonstrates that the knowledge and cognition of many of
our students is far lower than most people suspect, in spite of all that
has been written. Generally the people who are not shocked are those in learning
support services who spend a lot of time with students one-on-one, and really
know what students understand and what they don't.<br>
<br>
Jerry Epstein<br>
<br>
Cathy Kessel wrote:<br>
<blockquote type="cite"
cite="mid8C01164F-1249-4B95-A16A-5AE8E553ABB8@earthlink.net"><br>
<div>
<div>On Feb 10, 2006, at 4:04 PM, Jerome Epstein wrote:</div>
<br class="Apple-interchange-newline">
<blockquote type="cite">
<div style="margin: 0px;">I think Charles is correct to be very concerned
about online placement.</div>
<div style="margin: 0px; min-height: 14px;"><br>
</div>
<div style="margin: 0px;">I would point out that on my Basic Skills Diagnostic
Test (BSDT -- write<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">for info if you want), there is a quite strong
correlation between the<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">score on the first half and the SAT I. The
first half of the test is<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">non-algebraic and all of its questions should
be answerable by a<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">competent 8th or 9th grader. What is quite
stunning is the relative<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">scale of the correlation. If one says that
75% on the 8th grade level<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">test is minimally competent this correlates
to 620 on the SAT. Below 600<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">one finds rapidly increasing probability that
one will see serious<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">problems at elementary school level math, approaching
virtual certainty<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">at 500 (the national average). This has been
shown now at 3<span class="Apple-converted-space"> </span></div>
<div style="margin: 0px;">universities, and with many hundreds of students
at my university. <br>
</div>
</blockquote>
<div><br class="khtml-block-placeholder">
</div>
Does this mean that those universities have done a study? I'd be interested
in a reference.</div>
<div><br class="khtml-block-placeholder">
</div>
<div>The (old) SAT is well known (at least in testing circles) to underpredict
(in general, on average) women's undergraduate mathematics course grades
relative to those of men. Marcia Linn and I summarized studies about the
SAT (and other things, including possible legal consequences of its misuse)
in:</div>
<div><br class="khtml-block-placeholder">
</div>
<div class="Publications">C. Kessel and M. Linn. (1996). Grades or scores:
Predicting future college mathematics performance. <i>Educational Measurement:
Issues and Practice, 15</i><span style="">(4), 10–14, 38.</span></div>
<div class="Publications"> <O:P></O:P></div>
<div class="Publications">M. Linn and C. Kessel. (1996). Success in mathematics:
Increasing talent and gender diversity. In Alan Schoenfeld, Ed Dubinsky,
and James Kaput (Eds.), <i>Research in Collegiate Mathematics Education II
</i><span style="">(pp. 101-144). Providence, RI: American Mathematical
Society.</span></div>
<div><br class="khtml-block-placeholder">
</div>
<div>The SAT is not designed as a placement test. I haven't kept up with
what happens with the new one as regards underprediction of grades, but (again)
the new one is not designed as a placement test, so any math department that
wants to use it that way needs to validate that use or worry about possible
legal consequences.</div>
<div><br class="khtml-block-placeholder">
</div>
<div>--Cathy</div>
<div><br class="khtml-block-placeholder">
</div>
</blockquote>
<br>
</body>
</html>