Chapter 7 - Statistics for the non-statistician

Question 1

How are quantitative data stored in a data set?​

Answer 1

• Original unanalysed data are stored in datasets
• Each participant as a case and each measurement as a variable.
• Each numerical result is a score/value for one variable in a sample.

Question 2

How do descriptive statistics summarise the characteristics of a sample?​

Answer 2

By using demographics and outcome measures, which relate to treatments or other influential variables.

Question 3

What assumptions apply to the results?​

Answer 3

Equality:
- The assumptions include group comparability
Normality:
- The shape of the data
Linearity:
- The nature and direction of causality

Question 4

How sure can we be that results from the sample will generalise to the wider population?​

Answer 4

• Inferential statistics control and quantify uncertainty
• Though conclusions about the sample are precise, applying them to the population involves some uncertainty that must be managed to avoid false conclusions.

Question 5

How is the clinical importance of treatment effects decided?​

Answer 5

• Clinical importance depends on whether the observed effect is at least as large as the minimum important difference
• If it is beneficial without causing harm
• The effect must also justify the treatment effort, cost, and risks.

Question 6

How are treatment effects summarised from multiple studies of interventions in a systematic review? ​

Answer 6

Meta-analysis combines results from multiple quantitative studies, using tools like:
- forest plots to display effect sizes and CIs
- Assesses study diversity through heterogeneity statistics.