Biostats Topic 7

what are we testing during x2 goodness of fit test

observed data and expected frequencies from a random model

how many variables are in x2 goodness of fit test

1

what is the null hypothesis of x2 goodness of fit test

data match an expectation of a specific probability model

what are you testing during a fisher exact test

two categorical variables each with 2 categories and small sample size (2×2)

how many variables are in a fisher exact and a x2 contigency test

2

what is the null hypothesis for a fisher exact and a x2 contigency test

no association between variables (independence)

what are you testing in a x2 contigency test

inferred association between two categorical variables each with 2 or more categories

what does a x2 goodness of fit test compare

count data to a model stated in the null hypothesis

what does the test statistic measure

the discrepancy between the observed and expected frequencies

larger X2 means what

larger difference

what does the sum of the observed count need to be equal to

the expected count

what are the assumptions of a x2 goodness of fit test and a x2 contingency test

that you have a random season; none of the categories should have an expected counts of less than 1 (no zeros) and no more than 20% of the categories should have expected counts less than 5

what do you compare your test statistic to

the approximated theoretical X2 distribution

how do you find your degrees of freedom in the x2 goodness of fit

#categories -1

what does calculating the degrees of freedom do

helps you identify what null distribution to use

what is the degrees of freedom

the number of pieces of independent information available when calculating a test statistic

the expected frequencies must add up to what

100%

what is the critical value

the value of X2 where P-value = alpha (0.05)

if your test statistic is larger than your critical value then you

reject the null hypothesis

if you test statistic is smaller than your critical value then

fail to reject the null hypothesis

how do you find the expected count in a X2 contingency test

expected count = (row total x column total)/ table total

when is the only time that you use a fisher exact test

with a 2×2 table

what does the fisher exact test not have

the same assumptions of a x2 contingency test

when do you use the fisher exact test

if assumptions of x2 contingency test are violated (usually small n)