Statistics Definitions

Question 1

Sensitivity

Answer 1

a/(a+c) or TP/(TP+FN)

SNout

Question 2

Specificity

Answer 2

specificity= d / (b + d) or
specificity = TN / (FP + TN)
SPin

Question 3

False Positive Rate

Answer 3

false positive rate = b / (b + d)

Question 4

Positive Predictive Value

Answer 4

PPV = a / (a + b) or
PPV = TP / (TP + FP)

Question 5

Liklihood ratio

Answer 5

positive likelihood ratio = sensitivity / (1 - specificity)

Question 6

Relative Risk

Answer 6

Risk of developing diseae with people with KNOWN exposure

Question 7

Odds Ratio

Answer 7

Probability of having risk factor is one HAS disease

Question 8

Number Needed to Treat

Answer 8

1/Absolute risk reduction therefore ARR=1/NNT

Question 9

Power

Answer 9

power = 1 - (probability of a type-II, or beta error

an estimate of the probability a study will be able to detect a true effect of the intervention power

Question 10

Type II error (Beta)

Answer 10

detecting no difference when there is None or INCORRECTLY accepting the null hypothesis

Question 11

Type I error (alpha)

Answer 11

Incorrectly rejecting the null hypothesis, detecting a difference when there is NOT one

Question 12

Student t-test

Answer 12

Comparing means of parametric date (2 means)

Question 13

Man Whitney rank sum, Wilcox

Answer 13

Comparting NON parametic data, (2 means). That nmeans NON normally distributed

Question 14

Chi Squared Test

Answer 14

comparing proportions or categorical data

Question 15

P Value

Answer 15

Probability , under assumption of null hypothesis, of obtaining result equal too or more extreme

Question 16

Power

Answer 16

Ability of a study to detect a difference if once truly exisists

Question 17

Baeysian Analysis

Answer 17

An analysis that starts with a particular probability of an event (the prior probability) and incorporates new information to generate a revised probability (a posterior probability)