Lecture 23: Chance 1

Question 1

How do bigger sample sizes help to improve the accuracy of a study’s findings?

Answer 1

It will not tell you about the accuracy of a study’s findings.

However, a larger sample size increases the likelihood that the sample will be representative of the underlying population, reduces the width of the confidence intervals, and therefore increases the likelihood of getting more precise estimates of the parameter - (the true value of the measure in the population being investigated).

Question 2

Interpret these findings:

Suppose that a randomised controlled trial which examined the association between stretching before physical activity (compared to no stretching before physical activity) and muscle injury reported the following findings: RR = 0.64 (95% CI: 0.50-0.78).

Answer 2

Participants who stretched before physical activity were 0.64 times as likely to experience muscle injury compared to participants who did not stretch before physical activity.

We are 95% confident that the true relative risk lies between 0.50 and 0.78. As the confidence interval does not include the null value of 1, the finding is statistically significant. Chance is an unlikely explanation for this finding

Question 3

Prior to the study the researchers stated that the effect of stretching would be clinically important if it reduced muscle injury risk by 25% or more. RR= 0.64
(95 % CI: 0.50 - 0.78)

Were their findings clinically important and why/why not?

Answer 3

The findings of the study are possibly clinically important. Please see your tutorial five notes for more information on clinical importance.

Question 4

What are the three factors that must be considered for internal validity?

Answer 4

Chance
Bias
Confounding

Question 5

What is external validity?

Answer 5

Extent to which the study findings are applicable to a broader or different population

Also known as Generalisability

Question 6

What is sampling error?

Answer 6

If you repeatedly sampled randomly from the same source population, most of the time you would get a sample with a similar composition to the population you sampled from.

But some of the samples would be quite different just by chance.

This is called sampling error and is a form of random error.

Commonly just called chance.

Question 7

How can sample size effect sampling error?

Answer 7

Increasing the size of the study sample when randomly sampling:
- Reduces sample variability (standard deviation etc.)
- Increases likelihood of getting a representative sample
- Increases precision of parameter estimate

Question 8

Can we eliminate sampling error?

Answer 8

Can’t eliminate sampling error but can reduce with larger sample sizes

Question 9

How to measure influence of sampling error?

Answer 9

Confidence intervals and p values

Question 10

What is the template that is used for making a comment about 95% confidence intervals?

Answer 10

‘We are 95% confident that the true population value lies between the limits of the Confidence Interval’

Question 11

How can confidence intervals help with determining clinical significance?

Answer 11

Confidence intervals can help us decide whether the study
findings are clinically important by their width and position

Clinical importance is often a set value eg. (0.9 RR is significant) so using a confidence interval to gauge whether it is above or below this value.

Question 12

What is a parameter and how is this different to a point estimate?

Answer 12

Parameter:
The true value of the measure in the population that the study is trying to discover

Estimate:
The measure found in the study sample
Sometimes referred to as the point estimate