Hypothesis tests of Proportions

Question 1

null hypothesis

Answer 1

no difference between groups or no association/
true difference in parameter on treatment or on placebo =0

Question 2

alternative hypothesis

Answer 2

contradicts the null hypothesis
states that the true difference in parameter on treatment or on placebo is not 0

Question 3

p value

Answer 3

probability of observing a test statistic as, or more extreme than that observed by tour sample in hypothetical repetitions of the study assuming that the null hypothesis is true

Question 4

if p<0.05

Answer 4

statistically significant at 5% level
null hypothesis rejected

Question 5

is p>0.05

Answer 5

not statistically significant at 5% level
insufficient evidence to reject the null hypothesis

Question 6

type 1 statistical error

Answer 6

if null hypothesis rejected when it is true
probability of error = significance level

Question 7

type 2 statistical error

Answer 7

if the null hypothesis is not rejected when it is false
larger studies reduce risk of type 2

Question 8

what is chi squared 2x2 used for

Answer 8

to compare proportions between 2 independent groups

Question 9

hypothesis tested in chi 2x2

Answer 9

H0 = population proportion in group 1 equals population proportion in gp 2 OR no association between group variable and outcome variable in the population
H1=pop proportion ingp1 does not equal the pop proportion in group 2

Question 10

assumptions in chi sqared 2x2

Answer 10

only considered reliable if all expected values are greater than 5

Question 11

what is absolute risk reduction

Answer 11

difference in risk between the 2 treatment groups

Question 12

absolute risk reduction equation

Answer 12

risk in placebo group - risk in treatment group
=b/(b+d) - a/(a+c)

Question 13

what is relative risk

Answer 13

risk in the treatment divided by the placebo group
above 1 = increased treatment group compared with the placebo group

Question 14

what is the odds ratio

Answer 14

odds in the treatment group divided by the placebo group
=odds in treatment group/odds in placebo group

Question 15

chi squared test for r x c table
usage

Answer 15

to compare an outcome categorical variable with r categories between c independent groups

Question 16

chi sq rxc hypothesis tested

Answer 16

H0= no association between group variable and outcome variable

Question 17

assumptions in chi squared test for r x c

Answer 17

only considered reliable is most expected values are greater than 5 and all exceed 2

Question 18

usage of chi squared test for trend

Answer 18

to test whther a proportion increases or decreases as the categories of a group variable increase

Question 19

the main advantage for using the chi-squared test for trend

Answer 19

if there is a trend across categories the test is more powerful than standard chi squared test

Question 20

hypothesis tested in chi squared test for trend

Answer 20

H0= no association between the group variable and the outcome variable
H1=proportion with a given outcome increases or decreases as the category of the group variable increases