Lecture 13 - Experimental and Quasi-Experimental Designs

Question 1

What is a ‘true’ experiment

Answer 1

experiments with random allocation of participants to treatment conditions

Question 2

why do we randomly allocate participants to conditions?

Answer 2

it spreads any confounding variables EVENLY across conditions, rather than there being a SYSTEMATIC difference other than the treatment

so it helps us to isolate the effect of the treatment variable

Question 3

what does O and X mean when describing an experiment?

Answer 3

O means observation/test
X means a treatment

so OXO is obvservation/test, treatment, then a second observation/test

Question 4

What are three systematic approaches to controlling for nuiscance variables?

Answer 4

Blocking - experiment arranged in blocks with the known nuiscance variable held constant, e.g. experimenter

Matching

Counterbalancing (eg AB BA)

Question 5

What is a ‘latin square’

Answer 5

A way to work out counterbalancing conditions
e.g.
ABC
BCA
CBA