Experimental design

Created by

Pierre Gasly

Cards (27)

What are the two reasearch designs
Observational and manipulative
What type of research study is correlation
Observational
What type of research study is descrtiptive
Observational
What type of research sutdy is experimental
Manipulative
Describe an observational study
It looks for patterns and associations. But cannot prove causation.
Describe a manipulative study
The conditions of the participants are manipulated.
It tests both the cause and effect of the variables.
It aims to prove causation.
What are the main princciples of experimental design
Control, randomization and repetition.
What are the features of a good research question
It must be clear and precise.
Why would you choose an observational sutdy of a manipulative one
The data is easier to collect.
It is less likely to have unintended consequences.
What type of data is often used for observational studies
Categorical
What is the downside to manipulative studdies
The manipulations may be impossible or unethical to perform.
What is reverse causation
When the response variable affects the explanatory variable.
What is a confounding variable
This is something that affects both the explanatory and response variable.
This means that the response and explanatory variables have no effect on each other.
What often follows on from observational studies
Manipulative studies
Why do you use a control in studies
It allows your experiment to be comparative and establish whether the difference is just a random change over time.
It also removes the possibility of the placebo effect.
Why do you need to be able to replicate results
Because individuals vary
What are the types of manipulative studies
Randomised controlled trial [RCT].
Describe a randomised controlled trial
It uses hypothesis testing for causality and prediction.
The generalizability of the results is said to be sample dependent.
The question would be pharased as:
Does the intervention work for this sample?
What are the types of observational studies
Meta-analysis of randomised controlled trials.
Quasi-RCT known as clinical controlled trials.
Case-control study.
Cohort study.
Cross sectional study.
Simple case study known as a survey.
What type of research design has the highest generalisation
Meta analysis of randomised controlled trials
What type of research design has the lowest amount of generalisation
Single case study
Define the generalizability of a study 
This describes whether it can be applied to different people such as the wider population.
Describe a single case study
It does not examine causality or effects.
It has no generalizability as it looks at treating an individual.
Describe a cross-sectional study
It analyses the data in a population at a single point in time.
It has limited generalizability and no causality.
It aims to look for an association between two variables.
What is the alternative name for a survet
Cross-sectional study
Describe a cohort study
It looks for an association between the independent variable and a negative outcome, which is the dependent variable.
It develops a causality hypothesis and so has some generalizability.
Describe a case-control study
It compares two sets of parameters in each group of subjects.
It looks for an association between the parameters, meaning it has some generalisability.