Confidence Intervals and Descriptive Statistics

Question 1

Mean

Answer 1

The average

Question 2

Median

Answer 2

The middle number

Question 3

Mode

Answer 3

The number that appears most often in a set

Question 4

Range

Answer 4

Difference between the lowest and highest values

Question 5

Skewed

Answer 5

The data does not have a “normal” distribution

Question 6

Standard Deviation

Answer 6

Measure of how spread out numbers are. If distribution is approximately normal then about 68% of the data values are within one standard deviation of the mean and 95% are within two standard deviations

Question 7

Normal Distribution

Answer 7

When the mean = median = mode
Symmetry about the center
50% of the values are less than the mean and 50% are greater than the mean
Imagine a standard bell curve

Question 8

Confidence Intervals

Answer 8

• Prove the range of values within which the “true population” effect is likely to reside
• Most of the CI is 95%; some studies use 90% or 99%

Question 9

Confidence Intervals vs P-values

Answer 9

• Both CI and P-values help if the research results are the result of chance or represent a “true” outcome
• In addition the CI gives us information about the sample size and variation in the sample

Question 10

Absolute Risk Reduction (ARR) formula and how to apply this to the CI to determine if the results are statistically significant

Answer 10

Formula: Control Event Rate (CER) - Experimental Event Rate (EER)

The null hypothesis is that there is no difference (EER - EER = 0)

So if the 95% CI for a study about the ARR includes zero, the difference between EER and CER is not statistically significant

If you are using differences, then 0 = no statistical significance

If you are using ratios, then 1 = no statistical significance