Some things are easy to measure. Like a greazt kick. If it scores, it is obvious. However, in a match one good kick may not be the deciding factor. To be a success you have to decide what are ALL the factors that need to be measured.
To be a good instructor there is more than just the instructor's ability as a martial artist. Unfortunately many instructors are chosen because they stuck around longer than their colleagues or are a high rank. Neither of these factors are relevant in the search for a good instructor or in measuring an instructor's performance.
So what do we know? We know that instructor output (performance) affects student attendance and grading performance - true.
Now we need to measure the results:
1) Student attendance is based on how many times a student attends class. So, if 2 instructors both have 10 people attending their class, it looks like they are equal. But what if one instructor has 20 enrolled in their class and the other has 10? That means that one has 100% attendance and the other has 50%.
Our goal is 100% attendance. So measure the attendance percentage.
2) Quit rate is how many people quit on a monthly or yearly basis. Instructors directly impact this.
Work out how many students would quit in a good month. That would be 0.
Now how many is acceptable? 2 or 3? An industry standard is 5% or below, depending on the size of the school.
Then monitor the individual class quit rates on a monthly basis.
3) Grading standards is in two parts:
a) How many grade, and
b) The student's ability level.
a) is easy. If you have ascertained that it takes 3 months to move from one level to the next, then your class members should last 3 months in that phase of the curriculum. If the instructor takes longer, it may mean he or she is sub-standard or it may mean the students are struggling with the content (Student's fault or content too hard?)
b) Grading standard. I found this hard. Do we measure students against each other? Against other schools? How about a numerical score? None of these are fair and/or easy to measure.
In summary, I would start at the top and measure attendance first, then work my way down.
P.S. Email me here if you need some templates.