IV types vs levels
Types are nominal, Levels are ordinal
What types of scales permit ANOVAs and t-tests?
The ratio and interval scale
What types of scales permit Chi-square and McNemar’s tests?
Nominal and ordinal scales
Within participants design
Participants participate in both levels of IV
Between participants design
Participants only participate in one of the groups
Mixed designs
Participants seperated into groups (between) and do multiple levels of IV2 (within)
Pre-post design
Time based IV, within participants
Specifying IVs
IV#:
Name,
#-levels (A, B),
[PAV IV?]
between/within,
manipulated/measured
E.g. IV1: Colour, 3-levels (Red, Green, Blue), within-subjects, manipulated
E.g. IV2: Gender, 2-level (Male, Female), PAV IV, between-subjects, measured
E.g. Baseline/ Control: target absent trials
Specifying Designs
Levels of IV, (factorial), else non-factorial
within/ between/ mixed/ pre-post
E.g. 3 x 2 mixed factorial design
E.g. Non-factorial design
Specifying DVs
DV#: Name (Likert range)
E.g. DV1: Response time
E.g. DV2: Attractiveness rating (7-point scale, 1-7)
PAV stands for; examples
Person attribute variable
E.g. physical attributes (gender, height, weight), psychopathology (diagnoses?), opinions, livelihood (where they live, car they drive, etc.)
PAV IVs are dangers to what validity?
Internal
If no random assignment is possible, how can we determine whether the construct is isolated in the relationship?
Randomised control trials
RCT have random assignment into groups; new treatment, and either existing treatment, placebo, or no treatment (from best to worst, respectively)
