Week 6 - Research Design and Sampling

Question 1

Pseudo-random binary sequences (PRBS)

Answer 1

Use to reduce bias in random sampling

Question 2

Pretest + posttest

Answer 2

Use same group, baseline first then experiment/control group.

Question 3

Randomised posttest

Answer 3

Posttest only - no baseline

Question 4

Matched-pairs design

Answer 4

Match people that are similar and split them into different groups

Question 5

Repeated-measures designs

Answer 5

Uses same individuals over and over

Question 6

Statistical matching

Answer 6

Use instead of matched pairs for covariates

Question 7

Strengths of RCT

Answer 7

• shows significant effect
• low researcher subjectivity

Question 8

Limitations of RCT

Answer 8

• expensive
• need lots of ppl
• low external validity

Question 9

Quasi-experimental designs

Answer 9

At least 1 of 3 is missing:
1. random sampling
2. control groups
3. manipulation

Question 10

Non-RCT

Answer 10

Self-selected

Question 11

Interrupted time series

Answer 11

Observations - treatment occurs naturally or administered over time

Question 12

1 group posttest

Answer 12

Limitation = no baseline/control group

Question 13

History effect

Answer 13

External event biases DV

Question 14

Maturation effect

Answer 14

Change with time bias

Question 15

Instrument decay

Answer 15

Gradual loss of measurement accuracy

Question 16

Regression toward mean

Answer 16

If 1 is extreme then the next is closer to mean