4/18/12
Stastically Socialstat significant the results are representative of the population
Significance testing
if the probability is small that the difference in the sample occurred by chance ( which means
the probability is high that the diff. occurred because there is a diff. in the population) then
our results are statistically significant
if the probability is low that the difference in the sample occurred by chance ( which means
the probability is low that the diff. occurred because there is a diff. in the population) then our
results are not statistically significant
significance testing
hypothesis a statement about the relationship btw 2 variables
null hypothesis no relationship between the variables (no difference btw the groups and the
pop.)
alternative hypothesis there is a relationship between the variables (is a diff. btw groups and
pop.)
p value the probability that we observed the sample value given that the null hypothesis is
true
(observed sample value is the result of chance
Small value p means that it is not likely that our sample result is due to chance
Large p value means that it is likely that our sample result is due to chance
we want a small p value where it is not likely our results are due to chance and therefore
statistically significant enabling us to generalize to the population
compare p-value to an alpha level (alpha level is predetermined)
most common alpha level is .05 meaning only 5% of time null hypothesis is false when its
rlly true
if p-value is smaller than the alpha level, then our results are statistically significant at the
alpha level
other common alpha levels - .01; .001
if we see a difference in our sample and the results are statistically significant ( p-value is
smaller than the alpha level) then we reject the null hypothesis (basically accept the
alternative hypothesis)
if we see a difference in the sample but the results are not statistically significant ( p-value is
larger than the alpha level) then we fail to reject the null hypothesis