Metacognition D&M Ch3 Concept Review

Concept Review

What aspect of methodology distinguishes between the various metacognitive judgments -- JoL versus FoK versus Ease-Of-Learning?
This part of the Nelson model is called Metamemory Monitoring

Why are various metacognitive judgments linked more closely to different aspects of control? Why JoL --> self-paced study, FoK -->retrieval?
Judgment of Learning is linked to self-paced study because, given flexible time to study an item, the assumption is that people will typically study an item until their JoL is high. Feeling of Knowledge is linked to retrieval (rather than recall) because it is a measure of whether you think you would be able to recognize the answer. If you have a high FoK and then do indeed recognize it, your feeling was more accurate.

Compare Relative Accuracy and Calibration
The distinction between relative accuracy and calibration is that relative accuracy is a measure of your personal self-consistency -- this is important because in responding to questions like FoK and JoL, we are applying a mathematical scale to something we probably don't or maybe even can't internally measure mathematically. If someone is highly consistent, you can correct their numbers to smooth out the impact of personality/psychology due to their overconfidence/underconfidence. Measuring calibration is a way to see just how well attuned someone is to their actual condition internally -- are they overconfident, under confident, etc. Relative accuracy might be measured on a per-item basis (Yes, I think that one was right, or No, I'd like to come back to it later, etc.), while calibration based on an overall average at the end ("I think I got an A"). Might be really useful to look at SAT data where students can mark certain answers as ones to be reviewed later?

If JoL is measured on a 6 point scale, can you interpret relative accuracy? How about calibration?
To measure relative accuracy, you can use a 6-point scale, and assess their level of consistency that way (how sure you are on this item, versus how sure you are on that item, and how does that compare to your actual performance.. But calibration deals in direct deltas of accuracy between your judgment and your performance (percentage versus percentage), so you need to use the same numerical basis for the score.

How would calibration curve look for accurate, under confident, over confident?
Accurate is a 45 degree diagonal, under confident is over the line, over confident is under that line. This is because a typical graph places JoL on the X-axis and actual performance on the Y-Axis. Presumably they could be written the other way around, but the convention is to place the independent variable on the x-axis and the dependent variable on the y axis. So your recall is the facts-of-the-matter (did you get it right or not, presuming you'll do your best on the test of course) but your JoL is going to vary based on your skill at self-assessing.