It is not at all obvious. The main driver of debt level is the withdrawal of government support for higher education at both the state and federal levels, which has left universities to make up the deficit through increases in tuition.
And imposing a false interval scale on data that are at best ordinal doesn't improve the accuracy of measurement; it degrades it. Changing IQ scores to letter categories (say, at one sigma intervals) would be a good thing. Do you really think that a ten point IQ score difference is much of a predictor of anything, aside from maybe the score on a repeat test? On Feb 19, 2014, at 11:46 PM, Mike Wiliams wrote: > Given the level of education debt in the country, it's obvious that colleges > and Universities are making far more money than test companies. Has anyone > ever calculated how much information is lost by converting a perfectly good > test average into a letter? Did I say letter? We actually convert scores > into letters? Imagine if we converted IQ scores into letters. Does anyone > know the history of using letter grades? The error in grading as a > measurement device contributes to the lower predictive power of grades. If > we scored courses better, I am willing to bet that they would be completely > redundant with SATs etc and standardized testing would have no unique > predictive power. > > Mike Williams > > On 2/20/14 12:00 AM, Teaching in the Psychological Sciences (TIPS) digest > wrote: >> Assessment companies and the test prep companies that live symbiotically off >> of them make a great deal of money. The test score is held up and apart from >> the grades as being somehow more fair. So I think they invite the scrutiny. >> >> I think any individual grade from the student's middle school or high school >> record might be less useful than an aggregate GPA. The 20-30 instructors >> together make an index with considerable predictive power. Not that they >> shouldn't be held accountable also. But it's unlikely that all 20 or so are >> grading too easy or too hard. And no individual instructor has the same >> financial investment in his or her product than the handful of institutions >> making coin from theirs. >> >> That being said, SES, for both grades and test scores, is a problematic >> variable to tease out from merit/ability to succeed in higher education. >> >> Nancy Melucci >> Long Beach City College >> Long Beach CA >> -----Original Message----- >> From: Mike Wiliams<[email protected]> >> To: Teaching in the Psychological Sciences (TIPS)<[email protected]> >> Sent: Tue, Feb 18, 2014 11:10 pm >> Subject: Re:[tips] SAT and High School grade study >> >> These studies of SAT and grades as predictors or criterion just >> >> highlight how grades are poorly designed as a measurement device. What >> >> is their reliability and validity as measures of performance. Somehow >> >> the college board and SAT makers get the scrutiny that we don't apply to >> >> ourselves as grade makers. The error goes both ways. Paul Brandon Emeritus Professor of Psychology Minnesota State University, Mankato [email protected] --- You are currently subscribed to tips as: [email protected]. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=34414 or send a blank email to leave-34414-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
