Week 2 Experimental design and sampling Flashcards
What is the independent variable?
What the researcher is going to manipulate
What is the dependent variable?
What we measure, does it change?
What are the types of research design?
Non-experimental or observation Experimental
Give example of non-experimental or observation study design
Cross-sectional Longitudinal/cohort
What do researchers do in non-experimental or observation studyies?
Nothing is manipulated by the researcher. They simply analyse the natural differences in the IV against changes in the DV.
Give examples of experimental studies
Quasi-experimental True experimental Pre-experimental
What do the researchers do in experimental studies?
The IV is manipulated by the researcher. They test whether this causes the change in the DV
How can causality be achieved?
Cause and effect must correlate Cause should precede effect All other explanations (confounding variables) of the cause and effect relationship must be rule out.
Why are experimental designs potentially better at determining causality?
Control for confounding factors so rival hypotheses can be rejected Isolate the cause (IV)
Describe a pre-experimental design
No random assignment of groups No control group One group post-test or pre-post test
Describe a true experimental design
Controls most threats to validity Random assignment to groups Control groups present
Name the study design
Post-test only group design
Name the study design
Randomised pre- post-test group
(Randomised Controlled Trial- RCT)
Name the study design
Cross-over design
Name the study design
Solomon four-group design
Name the study design
Static group comparison
Name the group design
Randomised post-test group design with a categorical variable
Name the study design
Time series design
Name the study design
Reversal design
Why do we sample?
Draw conclusions about a wider population
What are the two types of sampling?
Probability sampling
Non-probability sampling
What is probability sampling?
Selecting from a population where each member has a chance of selection and the probability of selection is known
What is non-probability sampling?
Selecting a sample when the researcher does not know what the population size is or has access to it. Probability of selection is therefore unknown.
Give examples of probability sampling
Simple random sampling
Stratified random sampling
Systematic sampling
Cluster sampling
Give examples of non-probability sampling
Convenience sampling
Snowball sampling
Describe simple random sampling
Each and every member of the population has an equal chance of being selected for the sample.
Describe stratified random sampling
Involves stratifying (dividing) the population into gorups based on known characteristic (e.g. age, gender)
The sample is then drawn from each group to mirror the proportional representation of the group.
Describe systematic sampling
Experimenter selects from an ordered list every 5th or 10th etc from a random starting position
Descibe cluster sampling
The sample unit is not an individual person but a cluster of people belonging to a ‘naturally occuring’ group. e.g. sities, countires, blocks or buildings or schools
Describe convenience sampling
The samples are selected because they are accessible to the researcher
Describe snowball sampling
The researcher asks the initial participants to indentify another potenitial participant who also meets the criteria.
