Testing hypothesis Flashcards
what is point estimate?
A single value to estimate a population parameter from a sample of
estimation
it is subjected to sampling variation.
variance inversely proportional with the sample size. bigger sample size = smaller variance, smaller sample size = bigger variance.
what is confidence interval.
An interval that, with high probability, contain the unknown (true) parameter in
interest (e.g.
m, p)
what does it mean it the confidence interval is wide?
it is less precise
what does it mean if the confidence interval is narrow?
it is more precise
95% Confidence interval interpretation:
If the experiment/procedure
were repeated on multiple samples, the calculated confidence interval
(which would differ for each sample) would encompass the true
population parameter 95% of the time
What does it mean if the Confidence interval crosses zero/ crosses the line? eg. ( -3.5, 1)
means it is not significant
What does it mean IF the confidence interval of odds/ relative risk crosses 1? eg. 0.5, 1.2)
no effect
what is the value of Confidence interval to determine that there is no effect based on the odds ratio/ relative risk
1
is this a one tail test or two tail test? H0: m=50 verse H1: m≠50
2 tail
is this a one tail test or two tail test? H0: m=50 verse H1: m<50
1 tail test
is this left or right tailed?
H0: m=50 verse H1: m<50
(left-tailed test) testing reduction
is this left or right tailed?
H0: m=50 verse H1: m>50
(right-tailed test) testing for improvement
what does H0 mean?
the null hypothesis of the test (usually No difference, No correlation, Not
dependent, etc.)
what does H1 mean?
the alternative hypothesis of the test (usually With difference, With
correlation, Dependent, etc.)
what is type 1 error?
false positive
H0 is true but u rejected it and accepted H1.
No result but u say have.
