Statistics Lecture

Question 1

Nominal groups

Answer 1

No hierachy
Binary outcomes e.g. dead or alive

Question 2

Ordinal

Answer 2

Data that can be ranked
E.g. Strongly agree, Agree, Neutral

Question 3

Continuous data

Answer 3

Scales e.g. height, temperature

Question 4

Parametric data

Answer 4

Can be measured
Always interval, ratio data
Normally distributed (Bell Curve)

Question 5

Non-parametric

Answer 5

Ordinal or nominal data
Does not have to be normally distributed
Data can be skewed (only if massive sample size)

Question 6

Outcome Variable/ Dependent variable

Answer 6

Variable that is measured
e.g. RR before and after exercise

Question 7

Independent Variable

Answer 7

Variable changed by researcher
Peanut vs cashew in study about strength of allergic reaction

Question 8

When looking at which tests to use

Answer 8

Is it testing difference between data?
or
Testing for a relationship?

Question 9

Anova test (analysis of variance)

Answer 9

Difference
Continuous data (e.g. time)
3 different unrelated groups
Normal distribution

Question 10

Pearson’s Product Moment

Answer 10

Relationship
Interval/ ratio
Normal distribution

Question 11

R2

Answer 11

Value between 0-1
Tells you how much of a relationship exists between one variable and another
Expresses percentage e.g. 0.85 = 85%

Question 12

Paired T-Test

Answer 12

Same people having test before and afterwards
Normal distribution e.g. BMI

Question 13

Linear regression

Answer 13

Simple or Multiple factors
To make predictions

‘logistic’ if prediction outcome is binary

Question 14

Confidence interval can only be significant if

Answer 14

both values are on same side of zero
Cannot include zero

Question 15

Narrow CI

Answer 15

Stronger power
more values have been included

Question 16

Accepted level of P value

Answer 16

<0.05
Smaller the value the better

Question 17

Sensitivity

Answer 17

Proportion of people who have a disease, who will test positive

How good is test at ruling people in.

Question 18

Specificity

Answer 18

Proportion of people without the disease who test negative

How good is test at ruling people out

Question 19

Kaplan-Meier curve

Answer 19

Time till event curve or cumulative survival

e.g age breastfeeding stopped