wk 3 Flashcards
variance
different performance on our measure(DV)
treatment variance
differences due to what we’ve done (different conditions/treatments)
error variance
differences due to other variables/factors
ceiling effect
the task is too easy
floor effect
the task is too hard
Measurement sensitivity
The ability to detect a change if one occurs
How to fix extraneous variables due to participant characteristics
Randomly assigned participants to groups if it is a between participants design
How do extraneous variables differ from confounding variables
confounding variables differ systematically with aspects of the design and therefore may provide an alternative explanation of whether an affect is found or not
Order effects
changes in behaviour because of ‘when’ a certain condition is completed
Practice effects
participants likely to get better at a task each time they perform it
Fatigue effects
If there are a number of conditions participants may get tired or bored which will affect performance
habituation
Participants may become less sensitive to a stimulus through repetition
Latin square design
this is used when counterbalancing variables when there are multiple conditions. E.g. 4 conditions = 24 orders.
provides a practical compromise to pure counterbalancing and the threat of order effects
Practice trials
these are included to avoid contaminating the first real trials (aka experimental/ test trials)
Randomisation
presents the individual trials in a random order using a computer. this helps to eliminate any systematic bias