1.3 Research Design

Question 1

anecdotal evidence, why case study more meaningful, correlation coefficient, need for a control condition, true experiments, quasi-experiments, correlational studies (pros and cons), random allocation, random selection, external validity, internal validity, blindness, replication

Question 2

anecdotal evidence

Answer 2

interpreted stories about a single past occurrence, usually of no scientific value, small sample size, only relevant info (no context), theory only supported by mentioned, only mentioned to support (cycle, bias), change w/ each retelling, can at most lead us to consider possible future research

Question 3

case study

Answer 3

• more systematic than anecdotes
• small sample size
• unbiased recording of data = never irrevocably linked to an explanation
• can be added together to give insights which can be added together later
• often only way to study rare disorders + conditions

Question 4

correlational study

Answer 4

study with at least two variable from each case/person
* measurements just taken, NO CONTROL OVER ANY VARIABLE - just observation
* difficult to know direction of causation or if there even is
* INFERRING CAUSATION IS NOT POSSIBLE

Question 5

correlation coefficient

Answer 5

• sign = direction of slope/relation
• magnitude = how clustered the cases are along the line (i.e how strong the relationship is)

Question 6

control condition

Answer 6

not receiving the treatment etc - exists to rule out other causes of change
* first essential requirement of inferring causation

Question 7

random allocation

Answer 7

participants arrive at the study NOT BELONGING TO ANY LEVEL OF THE INDEPENDENT VARIABLE

Question 8

random selection

Answer 8

when participants are RANDOMLY CHOSEN FROM THE POPULATION

Question 9

true experiment

Answer 9

all independent variables of interest are controlled and able to be randomly allocated
* strong causal inference can be made
* if a difference is found in the dependent variable you can make strong conclusions (differences between groups cancel out)

Question 10

quasi-experiment

Answer 10

at least one key variable of interest cannot be randomly allocated, but others can
* e.g. study involving pre-existing groups - confounds which weaken causal inference
* most common type of research design in many fields of psychology

Question 11

external validity

Answer 11

extent to which findings from the study can be generalised to the population at large
* how was sample chosen? (random sampling, self-selection)
* where/how? (artificial, etc)

Question 12

internal validity

Answer 12

extent to which changes in dependent variable can be contributed to changes in the independent variable (true exp = high int val, corr = low int val)

Question 13

blindness

Answer 13

essential - blind participant = unaware of treatment, etc
blind researcher = random allocate correctly, correctly record data
* aim: double blind

Question 14

replication

Answer 14

• when same findings are found by an entirely independent party following the method you followed