Randomised Single-Factor Design Flashcards
What do single factor designs do?
investigate the effect of one factor on the dependant variable
How many levels does a single factor have?
2 or more
What is a confounding?
distortion of the association between the factor and DV
What does randomisation do?
minimises any other systematic differences between control and intervention groups
What is haphazard randomisation?
smaller animals may be better at avoiding being picked up
How can a questionnaire be efficiently randomised?
randomise who receives the questionnaire
randomise the order of the questions
randomise the stimuli within the questions
consider left-right, top-down reading bias
What is balanced design?
when all factors levels (treatment groups) have the same number of experimental units (participants)
Why is balanced design important?
tests will have superior statistical power and be less susceptible to homoscedasticity
What test do single factor designs use (parametric and non parametric)
parametric - T-test
non-parametric - Mann-Whitney Test
How is a study with one factor but more than 2 levels analysed?
One-way analysis of variance
Kruskal-Wallis test
Regression analysis
Why is comparing all variants to a common control favourable?
because the analysis would include the control and other variants with the same regression