You're looking at Less Wrong's discussion board. This includes all posts, including those that haven't been promoted to the front page yet. For more information, see About Less Wrong.

gjm comments on Tutoring Small Groups of Children (for money) - Less Wrong Discussion

7 Post author: bramflakes 20 February 2013 11:51AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Article Navigation

Comments (27)

Comments (27)

You are viewing a single comment's thread. Show more comments above.

Comment author: gjm 20 February 2013 10:24:25PM 1 point [-]

There's nothing wrong with password-guessing -- as long as you know when you're doing it and don't mistake it for actual understanding. I'd have thought a tutor could teach students those skills without ruining their ability to grab extra marks in examinations by password-guessing.

Comment author: Sarokrae 20 February 2013 11:22:02PM *  1 point [-]

I'd argue against this. I always saw through password-guessing as fake and not really understanding anything when I was young, but lacked the people skills to notice that the teachers and examiners wanted me to guess the password rather than demonstrate that I really understood (because I didn't understand why), and lost a few exam marks along the way to figuring that out!

Comment author: Decius 21 February 2013 12:00:03AM 0 points [-]

Password-guessing skills are the lowest hanging fruit in terms of improving grades; your experience seems to support that as well.

So-called "test skills", which improve performance on tests without improving mastery of the nominal subject of the test, are strong evidence of inefficiency in the school system. Are you proposing remaking the entire game of quittich instead of getting the bludgers better brooms?

Comment author: Sarokrae 21 February 2013 12:39:50AM *  2 points [-]

Sorry, I don't seem to have made myself clear. I was arguing against warning students against password guessing. I.e. don't remake the game, just play it as intended.

Comment author: Decius 21 February 2013 12:57:30AM 0 points [-]

There's a certain amount of remaking the game desired, but the way to remake the game isn't to tell students to follow the rules that should be in place instead of the rules that are in place.

What's the best way to teach password-guessing skills? Given a small number of mutually exclusive choices (as in a multiple choice or true/false exam), how do you determine the one that the creator of the question intended without knowing enough about the specific subject?

Comment author: faul_sname 21 February 2013 06:16:59AM 3 points [-]

Given a small number of mutually exclusive choices (as in a multiple choice or true/false exam), how do you determine the one that the creator of the question intended without knowing enough about the specific subject?

  • In general, the second highest numerical answer is right about half the time.
  • When there are all-of-the-above or none-of-the-above questions, the person writing the test will choose the all/none-of-the-above choice either very rarely or above 50% of the time. Thus, if you know that the answer to one all-of-the-above question is "all of the above" and have no clue on the current question, "all of the above" is probably your best guess.
  • Surprisingly frequently, wrong answers don't fit grammatically into fill-in-the-blanks type multiple choice. If it's not grammatically correct, it's probably not the right answer.
  • Read the entire test before you start answering questions. Frequently, answers to early questions are in later questions. Here it's a good idea to learn to read at least twice as quickly as the average student so as to be able to do this with any fair time limit.
  • If two options are the same, they're both wrong. If one is the negation of the other, one of those two is correct.
  • Always guess. "Guessing penalties" don't actually penalize you for guessing, they're designed to offset random guessing on average. Your guess is probably better than random.
  • Unless you can point out the specific way in which the answer you gave first was wrong, don't change it. It's very hard to think "this answer is probably wrong" without thinking "this alternative answer is probably right", and so while your current answer is likely wrong, your next choice is likely to be worse.
  • If an answer contains qualifiers (may, might, sometimes), it's more likely to be correct.
  • One strategy I have heard of but not personally used is to treat multiple-choice questions as a series of true-false questions and pick the "most true" answer.

You're not going to go from random chance to an A with these techniques, but you might go from random chance to a 70% or from a 70% to an 85% (my ballpark estimate is that half of the errors the average person makes on a test are avoidable with the use of heuristics like these).

Powered by Reddit Powered by Reddit