In measurement they talk about norm referenced (like a curve) vs. Criterion referenced tests. But I like to say "scratch a criterion and find a norm". The 90/80 will fly out the window if everyone flunked in which case the test is made easier next time even tho it LOOKS like a strict criterion type test.
If memory serves the 90/80 came from Scottish teachers teaching in the US in the early 1800s. Does anybody know more about its origins? I never understood its persistence other than habit. Sent from my BlackBerry® wireless device from U.S. Cellular -----Original Message----- From: "michael sylvester" <[email protected]> Date: Thu, 8 Oct 2009 10:03:09 To: Teaching in the Psychological Sciences (TIPS)<[email protected]> Subject: [tips] To curve or not to curve Most profs probably assign grades based on preset criteria,for example,A (90-100), B(80-89) etc. But some profs may prefer to grade on the curve based on the mean and standard deviation. Just curious re the pros and cons of those differential methods.Which of these two methods are least immune to grade inflation? I have known some profs who have been accused of dishing out A s like hot cakes,but they have used grading by criteria. On the other hand they have been adjuncts and teaching at community colleges.Would grading on the curve solve that problem? Is there evidence that adjuncts give more of the A grade than regular faculty? Is there evidence that community college profs dish out the A grade like hotcakes? Michael Sylvester,PhD Daytona Beach,Florida --- To make changes to your subscription contact: Bill Southerly ([email protected]) --- To make changes to your subscription contact: Bill Southerly ([email protected])
