Lecture 8: Rigour and Experimental Designs

Question 1

What is the main epistemology of Quantitative Research?

Answer 1

Objectivist: Observe groups in an unbiased way (internal validity) and generalise the findings (external validity)

Question 2

What are the three criteria for Rigorous Positivist Research?

Answer 2

Valid: Describes the objective reality
Generalisable: study reflects the population of interest
Reliable: repeatable- over time and between researchers

Question 3

Experiments and Rigour

Answer 3

Aim: Control, and keep all the conditions similar
Differences are limited to the IV and DV…

Therefore, experiments ought to be:

1. Valid - Whatever we observe is the reality of that sample
2. Reliable - Repeatable, over time and between researchers
3. Generalisable - elements of control make it harder to be sure that the same effect will be observed elsewhere

Question 4

Stratification:

Answer 4

Sampling across the possible systematic sources of variation. When we cannot control for things such as age or gender, what we can do is try to remove the potential effects of confounding by employing the same numbers of females/males, ages etc. in the sample

Question 5

IV

Answer 5

the issue investigated to see whether or not it has an effect (the thing you change)

Question 6

DV

Answer 6

The outcomes were interested in to see whether or not an effect happens (the thing that we measure)

Question 7

what is an RCT?

Answer 7

it is when two or more clinical treatment options are identified. These are the I.V- an ‘arm’ (experimental group) is identified for each option

The Primary outcomes identified and measured are the Dependant Variables.

Question 8

What is statistical Power

Answer 8

Power is the likelihood of detecting a difference of a certain size that exists between the groups. The more people sampled, the more power the study has.

Question 9

Block Randomisation

Answer 9

Is random- so should limit some of the confounding but it might not give a balanced sample if we can’t impose a control

Question 10

Parallel-groups RCT

Answer 10

The standard form of allocation, which allocates each participant to an arm, one at a time

Question 11

Cluster RCT

Answer 11

Allocates participant to a treatment based on affiliation e.g. Villages, schools etc.

Question 12

How are RCTs analysed? (4)

Answer 12

1. Comparing the means/ hypothesis test
2. Logistic Regressions
3. Analysis of Covariance
4. Survival Analysis