The Incentive Effects of Normalizing Grades

I am going to try and summarize what we know and create new talk pages for the areas we are still trying to figure out. Hopefully this will make an unwieldy topic digestible again. I apologize if, rather than increases access and collaboration, these changes hinder learning. The “diffs” button on the top right of the screen will let you see the previous work if you want to edit my changes out.

I didn't condense the talk pages. If no one takes that up, and folks appreciate the reorganization, I will do so later.

-- AdamCarlis - 19 Feb 2008

The Incentive Effects of Normalizing Grades

Does the presence of a grade normalization curve cause students to refrain from helping one another?

-- EbenMoglen - 18 Feb 2008

Are Students Helping One Another?

The consensus seems to be that “people help each other in situations where they feel safe that their grade will not be affected negatively and reserve help where they think it might.” However, there are many students who fall outside that norm.

There are a few who never help others (many of us have had little or no contact with these students)
There are others who help without considering the possible negative repercussions

What Motivates those that Help?

Those students who offer limited help are motivated by:

A perception that the recipient is a non-competitor
A belief that explaining to others helps them learn the material
A belief that, by working with a small group, they can maximize their own performance

Those students who offer broad help are motivated by:

A belief that grading is haphazard and so see no consequences to their assistance
Anger about the system and so see helping others as a way to manifest that anger
The perceived utility of building relationships
A belief that helping others will not affect their grade

Is the Curve an Contributing Influence on those who do not Help Others

Those who either refuse to help or have come across students who do not help attribute that behavior to a belief that such help may negatively affect their own performance. Other possibilities include:

Insecurity about one's own knowledge
Poor social skills

"Shifters" indicate that the curve (or at least grades) have some influence

Students are more likely to lend assistance during the beginning of the semester and their willingness decreases to a low point right before finals. This seems to indicate that students, as they begin to focus on grades, are less willing to be collaborative (at least with those students outside their study group).

The Curve Heightens the Competitive Atmosphere

There is much to compete over in law school (jobs, student group boards, moot court, letters of recommendation, a good locker). Grades symbolize yet another competition point. While other grading systems also force teachers to compare students' work, the curve highlights that component of the evaluation system in the eyes of students. Since competition is about winning or losing (even if you have a study group, it just becomes a team sport) and the curve breeds competition, the curve feels like it is about winning and losing. As such, people are less likely to support other teams.

Structural Impediments to Collaboration

The organization of study groups creates an "us" v. "them" mentality, irregardless of a curve, which plays a role in the unwillingness to help students outside of one's group.

Grades are the Real Problem

Perhaps the curve heightens anxiety and competition, but those factors are created by grades themselves. It is hard to separate out grades from the curve and so it is hard to know which contributes most to people's unwillingness to help others.

Conclusion

Generally speaking, classmates report that some students refrain from helping others. While it is hard to differentiate whether that behavior is caused by the curve, grades in general, or simply the uncomradly nature of some students, there is some evidence that the curve at least heightens this behavior.

Continuing Exploration

Building a BetterLawSchoolTalk

Should we KeepTheCurveTalk?

Should Law School be Teaching us CompetitionInPreparationForCareersTalk?

To add to the "talk" section:

Interestingly, an empirical example of a school instituting a new grading policy is Princeton University where the administration "expects" departments to limit A family grades to 35%. This policy aims to revert back to a "pre-grade inflation" era when grades sustained more or less the same statistical distribution. Prof. Reinhardt, a brilliantly witty economist, attributes most of grade inflation at Princeton to professors' efforts in the 1960s to save students from the Vietnam draft by giving high grades to get them into graduate school.

In response to the new policy, undergraduates shared many of the concerns expressed on this board: (1) effects on post-graduation plans and (2) competitive and uncooperative behavior among students. The faculties in the science departments, with the exception of the astrophysics and geology departments in which a mere handful of students enrolled, strongly supported the policy. Among the students, the resistance to the policy, heavy at first, languished primarily because the administration concealed its design until a week before the faculty vote. After the vote, we felt helpless.

From students' beliefs, not necessarily the 'reality,' the stakes in the grading system appear higher in law school than college. Law firms recruit more heavily and discriminate among students more blatantly, while at the same time law school students cluster around the same employees to compete for jobs. Unlike undergraduate education, the interests of law school students, though diverse, are more homogeneous with respect to (a) curricula (b) career prospects and (c) analytical skills. Also, at least in the first year, law school generally judges students' work on exams more than papers. These are all generalizations, but virtually nobody outside of law school has experienced the rigor of torts, contracts, and civil procedure in preparation for a career, and very few in law school know a damn about mathematics or even science and engineering.

Recognizing these institutional differences, I can say that, generally, implementing the Princeton "curve" had almost zero effect on both students' incentives to cooperate and their ability to get jobs. I for one did not start to exhibit competitive behavior. Plus, the policy has not reduced the appeal of applying as more and more high school seniors have in every successive year, usually a sign of institutional 'prestige.' My point is this: I do not know whether it is the curve itself which is inherently evil or the structure and formatting of the law school experience in general. The assumptions listed above, I believe, suggest vast differences between the two types of educations which may go at the heart of any competitive effects flowing from the curve. Thus, reforming the curve requires reforms beyond the curve.

-- JesseCreed - 20 Feb 2008

I've attempted to create a voting system that way we can actually see how much grades matter to people in the class. It will be interesting to see if the distribution follows a bell curve. Another possibility is that we'll all end up on the GUNNER end of the spectrum (we did all end up in Columbia for a reason) or if we all end up in the less competitive end of the spectrum (which would be in keeping with less competitive anecdotes so far). Give me feedback and let me know if this is a good idea or not. Also let me know if this works, I'm having trouble figuring out how to display results.

What grades would I be satisfied with?		Mostly A's with Some B's 0% (2) All B's 0% (1) Mostly B's with Some A's 0% (1)

Disclaimer: I don't know much about wiki but I believe I've made this a closed poll which means everyone can vote once but can change it, and set the voting to secret which means no one else can see your vote.

[Note: Julian is correct; as this poll is setup, votes are secret and everyone can only vote once. Anyone interested in finding out more should take a look at the VotePlugin page which has a number of examples of how to use various options and types of polls. -- IanSullivan 21 Feb 2008]

Also, I realize by using a voting scheme I am relying on possibly statistical, scientific, and magical analysis to get an answer to a question. I apologize to those I may offend.

-- JulianBaez - 20 Feb 2008

I hope copying Julian's code comes out correctly, but I was very curious to compare Julian's question to another one, which is, how you would have answered the question before first semester grades came out. Were expectations lowered?

I had survey code inserted in this comment, but erased it because it didn't seem to be coming out correctly and I couldn't find a Twiki page on creating a survey. Julian, where are the instructions?

-- MichaelBerkovits - 21 Feb 2008

Hey guys, I'm planning on doing a survey as part of my re-write, so if you could hold off doing any survey on grades or the curve for just a bit, I'd appreciate it. If you have questions you think I should include, let me know.

Also, if you want to do a survey on something other than the curve, Eben left instructions in AmandaHungerfordFirstPaper?

-- AmandaHungerford - 21 Feb 2008

GradesChoice?

aOVNr3t0sIsDXLigOX2DKw

What grades would I be satisfied with?=All B's

2008-01-20

GradesChoice?

KufnVGCYfuNXC36Z5q1JZQ?