RCT (Issues with design, sampling and random allocation)

Question 1

The design of RCTs?

Answer 1

1) Participants are selected for the trail
2) They will randomly allocated to either the intervention group or the control group
3) The presence of the outcome of each group is then measured (was the outcome present/ not present in each group?)

Question 2

What is an ‘attention arm’?

Answer 2

A group similar to the intervention group in a RCT. However, they do not receive the intervention

Question 3

What type of hypothesis is stated at the start of a RCT?

Answer 3

Null hypothesis = states that there will be no relationship between the variables
This is said to be TRUE at the start of the study and then must be disproved once we know the results of the study.

Question 4

What are the 2 types of variables?

Answer 4

IV = is the characteristic you alter
DV = is the characteristic you measure

Question 5

Why must the target population (that contains the people you want to sample) be defined very carefully?

Answer 5

1) You want to know who can be included in the sample (INCLUSION) and who cannot be included in the sample (EXCLUDED).
2) You want to know when the results of the trial are known, who these results can apply to.

Question 6

What should the sample be?

Answer 6

The sample should be representative of the whole target population so that the results of the trail can be generalised to the rest of the target population

Question 7

What is probability sampling?

Answer 7

A sample that is a unbiased sample of people that meet the inclusion criteria and that we know have a chance of being selected

Question 8

What is non-probability sampling?

Answer 8

A sample that is a non-random sample of people, that we don’t know have a chance of being selected or not

Question 9

What are the 4 different types of sampling?

Answer 9

1) Simple Random
2) Stratified Random
3) Cluster Random
4) Systematic Random

Question 10

Simple Random:

Answer 10

Random selection of everyone in the target population using a random number table

Question 11

Stratified Random

Answer 11

Separating people from the target population into groups according to characteristics (that are thought to be of interest in the trial) and then randomly allocating people from these groups

Question 12

Cluster Random

Answer 12

Random sampling of people from larger units (e.g. hospitals) whereby you may not know every member of the target population

Question 13

Systematic Random

Answer 13

(when you know the target population size) Random selection of people at predetermined intervals (e.g every 20th member)

Question 14

What are the 4 different factors that could affect the sample size?

Answer 14

1) Population Factors
2) Design Factors
3) Measurement issues
4) Practical issues

Question 15

What is a Power Calculation?

Answer 15

It is performed at the start of a RCT and it will calculate the sample size that is required to adequately power a trial in order to discover where there is a relationship between the variables or not

Question 16

3 Population factors that affect the sample size?

Answer 16

1) The similarity of members in the target population (more similar - smaller sample size)
2) How often you would expect the outcome to be seen (if it occurs rarely - larger sample size)
3) Expected attrition (People leaving the trial)

Question 17

Design factors that affect the sample size?

Answer 17

If there are more variables to be tested then a larger sample size is needed

Question 18

Measurement issues that affect the sample size?

Answer 18

If the outcome of the intervention is hard to measure then a larger sample size is needed

Question 19

2 practical issues?

Answer 19

1) Cost

2) Convenience of recruiting people for the trial

Question 20

What are the 3 methods that can be used to randomly allocate people from the sample into either the intervention group or the control group of the RCT?

Answer 20

1) Simple randomisation
2) Block randomisation
3) Stratified randomisation

Question 21

Randomly allocating people from the sample to the 2 trail arms using 1) Simple randomisation

Answer 21

Random selection of every member in the sample using a random number tables

Question 22

Randomly allocating people from the sample to the 2 trail arms using 2) Block randomisation

Answer 22

Randomly allocating people to the 2 trial arms one block at a time (so there are similar numbers of participants in each trial arm, at one time)

Question 23

Randomly allocating people from the sample to the 2 trail arms using 3) Stratified randomisation

Answer 23

Allocating people from the sample to different groups depending on characteristics (that is seen to important to the trial) and then randomly allocating people from each of the groups to the 2 trial arms.

Question 24

Double blind trial

Answer 24

Neither the participants or the researcher knows which group is the intervention group or the control group. This is achieved by using a placebo.

Question 25

Single blind trial

Answer 25

When either the research or the participants know which group is which