CS: Stats

Question 1

(1-sensitivity) / specificity=

Answer 1

likelihood ratio for a negative test result
How much the odds of the disease decrease when a test is negative

Question 2

Correlation Tests
parametric & non-parametric:

Answer 2

Correlation
Pearson’s coefficient = parametric
non-parametric: Spearman’s coefficient

Question 3

X values lie within 1 SD of the mean

Answer 3

68.3%

Question 4

X values lie within 2 SD of the mean

Answer 4

95.4% of values lie within 2 SD of the mean

Question 5

within 3 SD of the mean lie X of the sample values

Answer 5

99.7% of values

Question 6

Likelihood ratio for a positive test result

Answer 6

= sensitivity / (1 - specificity)

Question 7

Standard error of the mean =

Answer 7

standard deviation / square root (number of patients)
gets smaller as the sample size (n) increases
- accuracy with which a sample represents a population

Question 8

What is use to indicate
sufficient ability to detect a difference between the two groups if indeed, a difference truly exists.

Answer 8

Power = 1 - the probability of a type II error
chance of NOT making a type 2 error
power increases (i.e. becomes closer to 1), the probability that making a type II error (incorrectly failing to reject the null hypothesis) decreases

Question 9

What indicates
Probability of obtaining a result by chance at least as extreme as the one that was actually observed, assuming that the null hypothesis is true

Answer 9

P-value
equivalent to the risk of a type 1 error (rejecting the null hypothesis when it is in fact, true).

Question 10

(EER - CER) / CER=

Answer 10

Relative risk reduction =

Question 11

Absolute risk reduction =

Answer 11

Control event rate) - (Experimental event rate)

Question 12

In statistics what is a type 2 error?

Answer 12

accepting null hypothesis when it is false
Type II error is inversely related to the statistical power - higher the statistical power, the lower the probability of making a Type II error.

Question 13

What is Type I error

Answer 13

• Rejecting the null hypothesis when it’s actually true
  -false positive results: you conclude that the drug intervention improved symptoms when it actually didn’t
• If the p value< significance level, it means your results are statistically significant and consistent with the alternative hypothesis.

Question 14

Define Odds

Answer 14

Odds - remember a ratio of the # of people who incur a particular outcome to the # of people who do not incur the outcome

Question 15

compares the percentage/proportions of patients who improved following two different interventions

Answer 15

Chi-squared test

Question 16

Plot that shows publication bias in meta-analyses

Answer 16

Funnel plots are usually drawn with treatment effects on the horizontal axis and study size on the vertical axis.
An asymmetrical funnel = publication bias or a systematic difference between smaller and larger studies (‘small study effects’).