Statistics

Question 1

Describe the 6 factors required for a normal distribution curve

Answer 1

• Bell shaped
• symmetrical
• has equal values for mean, median and mode
• has a single central peak
• is continuous
• has values between -infinity and +infinity

Question 2

Describe three ways to assess the normality of a distribution

Answer 2

1. Informal review - visulisation
2. Probability plot produced by statistical package
3. statistical test - KS and SW

Question 3

Describe the different options available for transformations

Answer 3

• logarithmic
• square root
  -reciprocal
• cube root
• logit

Question 4

Describe what data you would use a logarithmic transformation on

Answer 4

• fairly skewed data
• data where the variance is proportional to the mean

Question 5

Describe what data you would use a square root transformation on

Answer 5

-Data that is slightly skewed
- counts

Question 6

Describe what data you would use a reciprocal transformation on

Answer 6

Highly skewed

Question 7

Describe what data you would use a cube root transformation on

Answer 7

Data relating to volumes

Question 8

Describe what data you would use a logit transformation on

Answer 8

Proportions

Question 9

Describe the assumptions for using the paired t-test

Answer 9

• the distribution of paired differences is normal
• the differences are independent from each other

Question 10

What are the assumptions for the unpaired t-test?

Answer 10

• the data is plausibly normally distributed
• the population variances (or SD) of the two groups are equal

Question 11

What are the four options you can do if you want to do an unpaired t-test but the variances are not equal?

Answer 11

1. Welch’s tests
2. non-parametric test
3. data transformation
4. Do not proceed

Question 12

What does regression analysis do?

Answer 12

Gives you information about the nature of relationships between two variables

Question 13

What does correlation do?

Answer 13

gives you information about the relationship between linear continuous variables

Question 14

What does Chi Squared analysis do?

Answer 14

Gives you information about the relationship between two categorical variables

Question 15

when is logistic regression used?

Answer 15

When only one variable is categorical

Question 16

What is the correlation coefficient?

Answer 16

Describes the strength of the association between two variables

Question 17

What is the Pearson product moment correlation coefficient (p)

Answer 17

Used for the strength of correlation between linear relationships

Question 18

What are some assumptions for simple linear regression?

Answer 18

• observations are independent
  -relationship must be linear
  -residuals must be normally distributed
• homoscedasticity: equal variance

Question 19

Describe the non-parametric tests used for two samples of data?

Answer 19

• Wilcoxon signed rank test - for dependent samples
• Mann-whitney test - for independent samples

Question 20

Describe the two non-parametric tests that can be used for testing the hypothesis about a mean of a single sample

Answer 20

• sign test
• Wilcoxon signed rank

Question 21

When is spearman’s correlation coefficent used?

Answer 21

• sample size is small
• neither X or Y is normally distributed
• at least one of the two variables is measured on an ordinal scale
• the relationship is non-linear

Question 22

what are the assumptions for Chi-squared test?

Answer 22

• 80% or more of the cells have an expected value >5
• all expected frequencies >1
• total sample size >20

Question 23

What does cross-sectional mean in terms of study design?

Answer 23

Observes the subjects once

Question 24

What does longitudinal mean in terms of study design?

Answer 24

observes subjects over time

Question 25

What is a cohort study?
Describe some strengths and weaknesses.

Answer 25

Follows disease free subjects over time.
It reduces recall bias. Subjects can be randomised.

It’s large and costly, and you risk losing subjects over time.

Question 26

What is a case control study?

Describe some strengths and weaknesses.

Answer 26

Studies diseased subjects retrospectively.
Looks back in time to compare exposures between cases and controls.

Subject to recall and selection bias.
Cheaper and better for rare diseases.

Question 27

What is a cross-over study?

Describe some strengths and weaknesses.

Answer 27

Each subject acts as their own control. E.g. inhalers sequentially or excema cream on each arm.

Fewer subjects needed as variation reduced.
Wash out period needed to prevent interference of results.
Not suitable for all trial designs.

Question 28

What is a parallel control group study?

Answer 28

Treatment and control groups observed at the same time.

More subjects needed as larger variation

Question 29

What is case-controlled group study?

Answer 29

It’s a parallel control group where the treatment and control groups are matched for factors

Question 30

What is a sequential study?

Answer 30

It’s a parallel control group where the study is stopped once a difference becomes apparent.