Sample size Power Random allocation

1
Sample sizePowerRandom allocation

• R.Raveendran

2
Why is sample size important?

• Validity
• Accuracy
• Finance
• Resources
• Ethics

3

• What factors will affect the size of the sample?
• Degree of difference
• Type I error
• Type II error
• Variation of data
• Drop out
• Non-compliance

4

• What methods can be used to determine the sample
  size?
• Arbitrary numbers
• From previous studies ?
• Nomograms tables !
• Formulas
• Computer programs

5

• How to calculate the sample size?
• Assess the difference expected (0.5 kg)
• Find out the SD of groups (0.4)
• Set the level of significance (alpha - 0.05)
• Set the beta level (0.02)
• Select the appropriate formula (unpaired t)
• Calculate the sample size
• u-v (diff/sd)/
• Give allowances for drop-outs non-compliance

6
Power Probability that a study can detect a
difference Priori power determination Power
1 beta Beta or type 2 error is the chances of
missing a difference (false negative rate)
Posteriori power calculation Why? How?
What is the implication?
7

• Random Allocation
• Each unit in a sample has equal chance to be
  assigned a treatment
• Simple
• Block
• Stratified
• Cluster

8
Thank you
9
Degree of difference
The minimum difference that is clinically or
practically important e.g. A drug reduces BP by 2
mm of Hg (120 to 118). Is it clinically
important? What about 4 mm of Hg? What about
6? 10? 20? 30? 40? Implication
Large difference needs small sample size
Small difference needs large sample size
10
Type I and II errors
Question Is the rice cooked? Possible Results
(Yes) - (No) True
True False False
11
Type I and II errors
Groups Group A Vs BQuestion Is there a
difference between groups? Possible Results
(Yes) - (No) True
True False False Type I
error False P Limit - 5 Type II error
False - P Limit - 20
12

• Power Calculation
• Why?
• To find out whether a negative result is TRUE or
  FALSE
• How?
• Using the formula or computer programs
• What information is needed?
• The difference (0.5)
• Alpha (0.05)
• Sample size and SD both groups (5, 5 0.29,
  0.33)
• Statistical test used (Unpaired t)
• Implication? No power redo the experiment / no
  diff.