Hypothesis Testing
Z-Scores - Probable Limits
➔what are critical values?
◆ the scores that mark the point at whichanything beyondis considered significant(p
<.05)
◆ in between the two critical values on each side are the values that fall within, meaning
not significant values (p>.05)
➔for a z-distribution, the critical values measure95% non-significant scores(the middle chunk)
◆ the tailsadd up to 5%
➔most times .05is the cut off, but can be .01 as well
◆ reason for this is basically cause some old white guys couldn’t agree and everyone
collectively said screw them and landed on .05 🙂
One-Tailed vs Two-Tailed Tests
➔one-tailedis used when thedirection of hypothesisis specified(i.e. more than or less than)
◆ critical value cut off:+/-1.96
● tail would hold5%of scores
➔two-tailed is used whendirection of hypothesis isnot specified(i.e. just says difference)
◆ critical value cut off:+/- 1.64
● each tail holds2.5%of scores, adding up to 5%
Null Hypothesis
➔the null hypothesis basically states thatrandomnessalone is responsible for the effect
being observed
◆ some differences can exist but only due to “randomness” and not a significant factor
➔the whole purpose of hypothesis testing is to see if the null can be rejected, basically saying
more than just randomness is causing the effect
◆ you can either fail to reject the null (no significant evidence pointing to no randomness)
or you reject the null (significant evidence to show it’s more than just randomness)
◆ BUTjust because you reject the null,doesn’t provethe alternative hypothesis
Hypothesis Testing (Z-tests)
➔test to either reject or fail to reject the null hypothesis
◆ seeing if the observed effect is caused by randomness or the observed variable
Z-Scores - Probable Limits
➔what are critical values?
◆ the scores that mark the point at whichanything beyondis considered significant(p
<.05)
◆ in between the two critical values on each side are the values that fall within, meaning
not significant values (p>.05)
➔for a z-distribution, the critical values measure95% non-significant scores(the middle chunk)
◆ the tailsadd up to 5%
➔most times .05is the cut off, but can be .01 as well
◆ reason for this is basically cause some old white guys couldn’t agree and everyone
collectively said screw them and landed on .05 🙂
One-Tailed vs Two-Tailed Tests
➔one-tailedis used when thedirection of hypothesisis specified(i.e. more than or less than)
◆ critical value cut off:+/-1.96
● tail would hold5%of scores
➔two-tailed is used whendirection of hypothesis isnot specified(i.e. just says difference)
◆ critical value cut off:+/- 1.64
● each tail holds2.5%of scores, adding up to 5%
Null Hypothesis
➔the null hypothesis basically states thatrandomnessalone is responsible for the effect
being observed
◆ some differences can exist but only due to “randomness” and not a significant factor
➔the whole purpose of hypothesis testing is to see if the null can be rejected, basically saying
more than just randomness is causing the effect
◆ you can either fail to reject the null (no significant evidence pointing to no randomness)
or you reject the null (significant evidence to show it’s more than just randomness)
◆ BUTjust because you reject the null,doesn’t provethe alternative hypothesis
Hypothesis Testing (Z-tests)
➔test to either reject or fail to reject the null hypothesis
◆ seeing if the observed effect is caused by randomness or the observed variable
