0
$\begingroup$

Consider I have two search engines, one (B) with Mean Average Precision higher than the other (B). Now, can this metric be enough to say that the first search engine A is better than the second one B?

If not, how would other methods such as hypothesis significance tests would help me find out which one is better?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

No!

These are 2 common methods to find out better one,

  1. Two-sample t-test
  2. Two-sample Z-test

after test, you get either of 3 results........1)precision of group 1 is better, 2)precision of group 2 is better, 3)precision of both groups did not differ significantly.

Improve this answer
$\endgroup$
3
  • $\begingroup$ can you discuss this more? what does two-sample mean? $\endgroup$
    – ramborambo
    Commented Apr 20, 2015 at 14:37
  • 1
    $\begingroup$ two-sample are the samples of 2 groups which you want to compare. $\endgroup$
    – Hemant Rupani
    Commented Apr 20, 2015 at 14:39
  • 1
    $\begingroup$ why are those two better? $\endgroup$
    – ramborambo
    Commented Apr 20, 2015 at 14:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.