L9: Repeated Measures Designs

Question 0

Data from Gravetter & Forzano

Answer 0

• the same scores produce very diff statistical outcomes if they are from the same or diff Ss
  • e.g. Repeated measures design produces highly sig effect, has more power than ind groups design

Question 1

Repeated measures/within-subjects designs

Answer 1

• use of one sample of Ss
• includes 2 or more conditions IV’s
• control extraneous variables
• design may be factorial

Question 2

Advantages of repeated measures design

Answer 2

• economy: n=36 Ss tested in 6 cond. vs n=216 in an ind groups design (with no. In cells 36)
• large individuals differences b/w Ss do not obscure diff b/w treatment conditions

Question 3

Disadvantages of repeated measures design

Answer 3

time-related factors
- history of maturation - may matter if treatments occur at long intervals, not likely if treatments occur in a single session < 1 hr

Carryover effects
- specific treatment effects, treatment 1 (after effects) affects responses in treatment 2 - e.g. Drug effects

Progressive error
- general change due to experience in study - practice (+ve), fatigue effects (-ve)

Question 4

Controlling unwanted effects

Answer 4

• decide which conditions needs to be blocked rather than randomized e.g. rate of presentation
  • performance is likely to be more stable if Ss see lists at the same rate for a block of trials
• counterbalancing - present rates in diff order to different Ss
• randomize order of tasks or items, also order of words within a list

Question 5

Repeated measures design: visual search

Answer 5

• how do ppl find a vis target in an array?
• what are the characteristics of the search process?
• experimental task:
  • can measure speed or accuracy, speed most common, RT to detect presence or absence of target
  • experimental manipulations:
  
  1. Presence/absence of target
  2. Set size: number of items in display - 1,5,10,20
  3. Type of target: single vs conjunction of features

Question 6

An experimentation on repetition blindness

Answer 6

• occurs when ppl fail to detect and report a 2nd item in a briefly presented short of sequence of items
• subjects see some irrelevant items (digits) and report the relevant ones (letters) - each item shown for 1/10th sec.