Module 6

Question 1

What happens when we get more degrees of freedom?

Answer 1

• less variability
• closer to a normal distribution

Question 2

What sample size is generally safe?

Answer 2

n >/= 30

Question 3

What sample size is generally safe for highly skewed populations?

Answer 3

n >/=50

Question 4

What sample size is safe if a sample is not normally distributed but is highly symmetrical?

Answer 4

n>/=15 is fine for this.

Question 5

When determining sample size, what do we do if we don’t have sigma?

Answer 5

○ Estimate from previous studies
○ Use a pilot study to select a preliminary sample and calculate SD, and use that as an estimate for sigma
○ Use the range divided by 4
-(Largest - smallest)/4

Question 6

What is p*?

Answer 6

Planning value of pbar

Question 7

How do we get p*?

Answer 7

• Pbar from previous sample of same or similar units
• Pilot study pbar
• Best guess
• Or use p* = 50