Please enable JavaScript.
Coggle requires JavaScript to display documents.
Sample Size Calculation in epidemiology in studies (9, element determine…
Sample Size Calculation in epidemiology in studies
Objectives
17, randomized controlled trial (RCT)
Sample size calculation for RCT
Background
What is sample size
Population vs sample
Why we need to calculate sample size
Central limit theorem
What do you think about the use of the rukenof N ;30 in clinical study
7, factors that need to be determine and their impact on sample size.
8, steps in estimating sample size
Identify major study variable
2, determine type of estimate (mean ratio)
Indicate expected frequency of factor of interest
4, decide the desired precision of the estimate
5, decide the acceptable risk tnat esyimate will fall outsides its real population.
6, adjust population size
7, adjust for estimated design effevt
8, adjust for expected response rate
9, element determine the sample size
Error type 1
And tyoe 2
Z score and confidence interval
Variance
Precision
10, null hypothesis and alternative hypothesis
Drug X and drug Y shows no different
11, 2 types of error
Power of study (1 -B)
Power to reject the false null hypothesis
Power of the test
The probability of succesfully rejecting false null hypothedis
Ex. Type II error 0.2 or 20%, so, power of tne studh is 80 %
2 types of error
False pisitive
Alpha error (type 1 error) Allow 0.05 or 5%
False negetive
False negetive beta error type II error Allow 0.2 or 20 %
Type I error A
Probability of wrongly rejecting the HO when it js true
12, 1 sided or 2 sided tests
Example
1, 5 : 5 , ho is equal to 5 , ha not equal to 5 : 2sided hypithesis test because we look at equal or not equal .not equal may mean more or less
Ho less than equal 5 , ha more than 5 : one sided hypothesis test , because it is testing whether the mean is more than 5 or not.
More than equal of 5 , ha less than 5 means one sided because this test is mean less than 5 mmol or not.
13, precision
14, calculate sample size
Èstimating a proportion
Estimate duration of excercise
15, estimating a mean
The value of study can be judged by the width of confidence interval
15, sample size determination in various tupe of epidemiological studies
Descriptive studies
Case studies
Case series
Surveys
Descriptive ecologic studies
16, for epid studies (cohort/ case control studies)
Case conttol studies
Sample size calculation for case conttol studies
Cohort studies
Sample size calculation for cohort studies
18, non statistical consideration in determining sample size
Funding
Ethical issue
Number of patient availability
Novelty of the stufy
Similar study being underway
Method used to calculate sample size
STATA
SAS
SPSS
POWER AND SAMPLE SIZE CALCULATION
EPI INFO