[this applies particularly to work answers to take-home exams]

In general: 

see the related post on my blog from November 2015

Grades are one way of saying something about how I evaluate a student's work.  In that sense the grade is part of a conversation I am having both with the student and my colleagues, given the institutional context at Teachers College, a graduate school where all course work is more or less immediately related to a doctoral career.

It is also important to mention that a grade only reflects my personal reaction to a piece of work produced at a particular time and for a particular purpose. A grade for a paper or course does not necessarily reflect my overall evaluation of a student's career.

Specifically:

Most exam questions as I write them have a relatively "easy" answer if one has read the relevant required reading. This is the kind of answer that one might get if one asked it to current versions of "Artificial Intelligence" (e.g. ChatGPT). I will grade such an answer a B.

A better answer includes specific quotes from the sources mentioned (rather than just a summary of the argument made in the sources). I might grade this extra evidence a B+.

To reach an A, I expect to see evidence of original thinking that starts with the sources, with the relevant quotes, and suggests alternate responses to the overall subject matter. This must be more than a simple evaluation. It must include evidence of the thinking involved in making the evaluation or suggestion for further research.

In any event, B+ and above reflect my sense that the student should be successful in a doctoral career whether current (for students in doctoral programs) or in the future (for students in Masters program).

The higher grades reflect my sense that the student's work is either contributing something new or close to it.  The lower grades reflect my sense that the work is acceptable but that it has not yet reached the stage when integration has quite taken place.  Many successful doctoral students (and perhaps faculty members) never quite reach this stage, though we all try to do so.

It is quite common for me to hesitate between two grades.  In many cases, particularly in mid-terms and and other transitory papers, I will not try to resolve the ambiguity and will give two grades, one of which is indicated in parenthesis as the other grade that I also considered.  For example, some work might get a "B+ (A-)."  This would be an indication that the student is close to reaching a more interesting understanding of the topic.  A "B+ (B)" indicates that I am willing to give the benefit of my doubt but that I do have doubts.

 Marginallia:

A straight line in the margin is equivalent to a "good point," "that's right," and other positive remarks.  A wavy line is the reverse.

Finally

I am always glad to talk about the work and where a student might take it further. I am less inclined to discuss a grade per se.