So this semester as a teaching assistant, I've decided that I simply do not give out 100%s on lab reports. On quizzes and homeworks I'm happy to give out perfect scores, but on something like a lab report, I have determined that nobody is perfect and there is always something that could have been done better. As a concrete example, I could find nothing obvious wrong with a lab report last week, but his conclusion was a bit too long-winded and some of the stuff presented there should have been in the analysis, so I took off a point. He got a 99%. I feel that if somebody received a 100%, they stop trying to improve and perhaps get complacent. If somebody gets a 99%, they have a better chance at trying to figure out ways to improve. Meanwhile, the 1 percentage point difference on their lab grade is exceedingly unlikely to affect their letter grade, particularly if they're good enough to be getting 99%s on a regular basis. Does anybody have any thoughts about this? I can see the argument that perhaps it's unfair to take a point off from one person's paper that I don't take off from a paper that got, say, an 80%. But in my mind, the person who got an 80% has enough major stuff to work on without getting nitpicky. Are there any other arguments against this? Will this actual reduce the morale of higher-achieving students? Is anybody aware of any physics education research about this tactic? Has anybody used this to good effect?