Here's a different context for essentially the same question:
When I calculate my average FCI gain, do I...
(#1) calculate individual gains for each student, and then average the
results, or
(#2) do I calculate a gain using the average pre-test score and the average
post-test score?
Assuming not everyone is present for both tests, Method #1 involves a
smaller sampling size, although Method #2 is probably more rigorous. Even
with all the same students for both tests, though, you can see differences
in the <g>.
Here's a case where the choice is not based on "the underlying physics" of
the situation. Or is it?
_______________________________________________
Phys-L mailing list
Phys-L@electron.physics.buffalo.edu https://www.physics.buffalo.edu/mailman/listinfo/phys-l