[DanielStraight]

I can propose a few:

Have a one-on-one conference with each student where you ask them about their paper without letting them look at it.

Skip papers altogether and simply do a one-on-one interview about the topic to evaluate their knowledge of it.

Make students write papers in class.

Don't assign homework. Grade entirely based on class participation.

[steve-howard]

The last one will be extremely effective in deterring cheaters. It will also be extremely effective in deterring people who learn in different ways; every class I've ever had with heavy participation components was dominated by grade-obsessed people who spoke as quickly as they could. The end result is that reflecting on your response to a discussion means you won't be able to share it.

[Luyt]

While these are good ideas, I don't think someone would put them in practice. It'd make that person a very unpopular professor indeed, which will reflect in the student evaluations, and he can say goodbye to his yearly salary increase, and expect criticism from the dean about student and parent complaints about below average grades.

[derleth]

> Skip papers altogether and simply do a one-on-one interview about the topic to evaluate their knowledge of it.

OK, so you have to deal with the fact some people are intensely nervous around authority figures and a one-on-one is not going to reflect their knowledge: They'll be so nervous they won't be able to answer intelligently, or at their full capacity.

> Make students write papers in class.

Then for God's sake scale the paper sizes and complexity to class length. (You don't know how to do that. You think you do but you don't.)

> Don't assign homework. Grade entirely based on class participation.

This has all the downsides of the one-on-one coupled with all the downsides of large group dynamics: Some people are going to dominate the discussion, some people are not going to be able to get a word in edgewise, and there's nothing you can do to change that. The ones who dominate the discussion are not always the ones who are getting the material, and being a wallflower does not correlate with much of anything relevant to most college courses. (There are exceptions.)

[DanielStraight]

I realize there are weaknesses in all those suggestions. There are pros and cons to every grading method.

I think some of these can be mitigated though (whereas cheating probably can't, especially in my mind since I intensely disapprove of Turnitin).

A few ideas:

For the interviews, start by requiring office visits for more trivial things and build up to the main interview.

I had one professor that regularly started class by having every write for a few minutes (like 5-10) about something they found interesting in the assigned reading. Aside from making sure students actually read, assigning an amount of time rather than a topic allows students to select a complexity level they feel is appropriate for the time allowed.

For class participation, this obviously requires some re-thinking of what class participation means. It also probably requires smaller classes. In a class of 10 students, I find it unimaginable that the professor wouldn't know before grading anything approximately what grade each student will receive.