experimental design

Question 1

list the experimental designs

Answer 1

• independent groups
• repeated measures
• matched pairs

Question 2

define independent groups design

Answer 2

• 2 separate groups of participants experience 2 separate conditions
• eg. control condition & experimental condition
• performance of groups compared

Question 3

define repeated measures design

Answer 3

• all participants experience both conditions (A and B)
• 2 mean scores from both conditions compared to see if difference

Question 4

define matched pairs design

Answer 4

• participants paired together on variable/variables relevant to experiment eg. memory study (IQs)
• 1 participant from each pair allocated different condition
• attempt to control confounding variable of participant variables

Question 5

evaluate independent groups design (AO3)

Answer 5

-)
P: participants in different groups not same in terms of participant variables
E: if researcher finds mean difference between on DV this may be to do with participant variables than effects of IV. these differences act as confounding variable, reducing validity of findings

-)
P: less economical than repeated measures as each participant contributes single result
E: twice as many people needed to produce equivalent data & this increases money/time spent recruiting participants

+)
P: order effects aren’t a problem (problem for repeated measures)
E: participants less likely to guess aims

Question 6

how do researchers deal with the problem of participant variables confounding the results of independent groups design

Answer 6

random allocation

Question 7

evaluate repeated measures design (AO3)

Answer 7

-)
P: order effects
E: order could be significant or repeating two tasks can create boredom & fatigue. this may cause deteriorating performance on second task (eg. energy drink experiment - having energy drink first and water second may still have continuing effect of energy drink). or, performance may improve on skill-based task through practice. order acts as a confounding variable

-)
P: likely that participants will work out aim of study when experience all conditions
E: demand characteristics arise

+)
P/E: pariticpant varibales are controlled (higher validity) & fewer participants are needed (less time spent recruiting them)

Question 8

how are order effects dealt with in repeated measures

Answer 8

counterbalancing

Question 9

evaluate matched pairs design (AO3)

Answer 9

+)
P/E: only take part in one condition so order effects & demand characteristics less of a problem

-)
P: participants can never be matched exactly
E: even when identical twins used, there will still be important differences which may act as confounding variable

-)
P: matching may be time-consuming & experience (especially if pre-test required)
E: less economical than other designs