Types of research; experimental
- Created by: becky_99
- Created on: 09-12-19 01:31
Experimental research
Establish cause and effect:
- Cause must precede effect in time
- Cause and effect must be correlated
- Can do 1 study or multiple to try and establish cause and effect
Establish via:
- Selection of good theoretical framework
- Application of experimental design
- Use of correct statistical model and analysis
- Control of IVs
- Measurement of DVs
- Correct interpretation of results
Internal vs. external validity
Internal:
- Did IV change DV?
- Extent to which the results can be attributed to treatments used in the study
- Ability to generalise results
- Lab studies
External:
- To what populations, settings or treatments can the outcome be generalised?
- Field studies
Threats to interval validity
History - events that are not part of treatment.
Maturation - events due to passage of time.
Testing - effects of more than 1 test.
Instrumentation - change in calibration of measurements.
Statistical regression - selection based on extreme score.
Selection bias - non-random participant selection.
Experimental mortality - differential loss of participants.
Selection-maturation interaction - passage of time influencing groups differently.
Expectancy - experimenter's expectancy/anticipation of participants.
Controlling threats to internal validity
Randomisation - allows the assumption that the groups do not differ at the beginning of the experiment.
Blind set-ups - method where participant doesn't know whether they are receiving the treatment or not.
Double-blind set-ups - method where both participant and experimenter don't know which treatment the participant is receiving.
Placebos - control group receives a false treatment and experimental group receives real treatment.
Threats to external validity
Reactive of interactive effects of testing - pre-test may make participants sensitive to treatment.
Interaction of selection of biases and treatment - treatment may work only on participants selected on specific characteristics.
Reactive effects of experimental arrangements - setting constraints may influence generalisability.
Multiple-treatment interference - one treatment may influence the next treatment.
Controlling threats to external validity
Remove pre-test.
Represent a larger population.
- Participants
- Treatments
- Situations/settings
- Tests
Multiple experimental groups.
What is a true experiment?
A true experiment is where there has been manipulation of the IV and where units have been randomly assigned to the different levels of the IV.
Types of research design
One-shot study:
T O
One-group pre-test post-test:
O1 T O2
Static group comparison:
T O1
O2
(T - treatment; Blank - a control; O - observation; R - random allocation to groups; each line represents a group of subjects).
Types of research design; true experimental
Randomised groups design:
R T O1
R O2
Extending levels:
R T1 O1
R T2 O2
R O3
Soloman's 4 group design: only experimental design that controls all threats to internal validity, but difficult to analyse statistically.
Soloman's 4 group design
R O1 T O2 (group 1)
R O3 O4 (group 2)
R T O5 (group 3)
R O6 (group 4)
Causal-comparative vs. experimental research
ER - IV is manipulated by research.
CC - IV has already occurred.
ER - completely random groups.
CC - assignment to groups is based on pre-existing characteristics.
CC vs. ER approach
Research question example: does aerobic training improve VO2 max?
CC approach: measure VO2 max in marathon runners and sedentary healthy controls and compare the results.
ER approach: randomly assign sedentary people to an aerobic training group for 12 weeks or 12 weeks of habitual physical activity; measure VO2 max before and after.
Causal-comparative issues
Third variable - common cause.
Directionality - reverse causality.
Conditions necessary for establishing cause & effe
Statistical relationship between IV and DV has been established:
- ER = Yes
- CC = Yes
X preceded Y in time:
- ER = Yes
- CC = ?
Other factors did not determine Y:
- ER = YES
- CC = ?
Because control over possible extraneous variables is partial in CC research, some people think it can be very misleading and shouldn't be used.
Reliability and validity
Reliability: degree of consistency with which a measuring instrument measures whatever it's measuring.
Validity: degree to which a measuring instrument measures what it is meant to measure.
Reliability & validity in action within sport scie
VO2 max (gold standard fitness test):
- Expensive
- Time consuming
- Stressful for participant
Siconolfi's 3-minute step test:
- Cheap
- Easy to administer
- Not physically demanding for participants
Related discussions on The Student Room
- Visa type for visiting Ph.D. researcher »
- Becoming an astronaut with Cambridge degree »
- Experimental Psychology at Oxford Uni »
- HELP! University dissertation results, I tried but nothing »
- Online MSc Data Science »
- Best International Postgraduate Study »
- Manchester housing questionnaire »
- Struggling to decide EPQ question!!! »
- work experience/supercircular help needed!! »
- Which engineers use university-level education regularly? »
Comments
No comments have yet been made