<p>So are the higher level (200+) biology courses graded on a curve?</p>
<p>There are about 65-80 people in each class. some classes in particular- biol 221, 230, etc.</p>
<p>thanks</p>
<p>yes, all of those courses are graded on a curve...at least they were when i took them 2 years ago...everyone always thinks that there is no curve but there always is</p>
<p>yup, when i took it last semester, 202 was on a curve</p>
<p>ok thanks guys</p>
<p>Speaking of curves, why do professors tell you the standard deviation of scores on a test? Is the mean + 1 SD the approximate cutoff for an A or something?</p>
<p>yeah what's the standard deviation about? and is the mean usually a C or a B at Penn?</p>
<p>mean is set to about a c+ / b- (generally)</p>
<p>+1 sd is the a- to a range depending on the class</p>
<p>hmmm....</p>
<p>Ok so the mean score was a 60%. Standard deviation was 15%. </p>
<p>If we're a bit higher than the mean, what would the score translate to?</p>
<p>Say someone scores a 65% or a 70%. this is 1/3 SD and 2/3 SD over the mean. What grades would these receive?</p>
<p>And in Biology 121 (i heard this is a weeder class), is the mean set to a C+ or a B-? Does anyone know for sure?</p>
<p>If a class has more than 20 people, you can bet there's some kind of a curve that the professor uses to apply grades. The central limit theorem suggests that the distribution of scores in these cases approaches the normal distribution - this means that a professor can give a test of any difficulty, not worry about making a 90% score on the test be an A(for example) and use the scores produced to build the grade distribution. </p>
<p>Important in this process are the mean and standard deviation. The mean in this case is roughly the 50th percentile: half the class performed better and half the class performed worse. Std deviation? 68% of the scores in a normal distribution are within one standard deviation of the mean (if you score 1 SD above the mean, you've scored higher than 84% of the class). Professors often take this 84th percentile mark to be a B+/A- (the way they create this mapping is really up to the professor). </p>
<p>When the professor tells you the SD they're telling you either (for large classes) the a-/b+ cutoff range, or (for smaller classes, anything below 100) how spread out the grades were. A small std deviation means that the scores were all about the same - the professor did a good job teaching everyone but the grades will be hard to assign. </p>
<p>I realize this is more than most people are looking for, but it helps to understand what a professor is doing when analyzing test scores and assigning grades. Most professors don't have the time/ability to create perfect tests where 90% is an A (unlike the SAT/ACT, professors do not obsess over how the test is designed). These statistics help professors make meaningful grades from scores.</p>
<p>So how exactly are grades distributed? Tenebrousfire suggests a mean about c+/b- - this would be an ideal setup, but I think most of Penn has grade inflation that puts the mean higher (A recent DP article said that 54% of grades in the college were an A- or higher. This is not right) but it's difficult to make conclusions there.</p>
<p>One area where people get worried about grades is in upper-level classes that are very difficult. If professors took scores at face value, most people would fail. The 'curve' helps produce usual grades in these cases. </p>
<p>Another area is where scores are all tightly distributed and a professor applies a curve. In this case, a single question (or fraction of a point) can mean the difference between a C/B/A. This is a byproduct of closely-packed scores and an attempt to apply normal curve processes. If you feel this is the case, you may want to speak to the professor about his grading system.</p>