<p>Okay, so if a professor assigns a grade based on a non-normal distribution (such as an exponential distribution), how does it work out?</p>
<p>I know that when professors assign grades based on normal distributions, they calculate the z-score of a particular grade, and then a grade corresponds to the particular z-score.</p>
<p>Are there other distributions that are used when there are, say, a lot of very high (near perfect) scores, with a number of lower scores that are sparsely distributed?</p>
<p>From what I know, if even just one person gets nearly 100% on the exam while everyone fails, the prof won’t curve the exam, yes even if the whole class average is like 51%, that happened in one of the sections in my 1st year physics class. One person got 90% and everyone got likt 50%.</p>
<p>So will have a set percentage for each grade (often dictated by the department). </p>
<p>For example, one will give 20% A’s, 35% B’s, 40% C’s, and the rest D’s and F’s…</p>
<p>This rarely happens though… I’ve never been in a class where there has been an exponential distribution of scores.
I’ve had classes where there was a high normal distribution where like everyone was between 80 - 100 with like a 92 average, but nothing exponential.</p>
<p>^Depends on the prof, at my CC if it’s that bad, the professor ignores the 90%/gives them extra credit and curves it to the 2nd highest grade. Although some are hard-asses about it.</p>