lab experiments Flashcards
what are lab experiments
independent variables are isolated
controlled settings
clear designs
establishing causality is generally straightforward
what are the three key components of lab studies?
experimental manipulation
standardisation
random assignment
what is operationalisation?
how we define the variables we manipulate and measure
what is a pilot study?
small scale experiment to check that the IV works as intended, and that the experimental design is tolerated by PPs
what is standardisation of procedures
every PP experiences the experiment in the same way, barring differences between conditions
reduces extraneous variables
what are the key designs?
between-subjects = PPs only complete one condition each
within-subjects = all PPs complete all conditions (counterbalancing is important)
the type of design used dictates which statistical test you use to analyse results
what are the +/- of between subjects design?
+ PPs kept unaware of difference between conditions
+ no chance of order effects
- different groups may have different characteristics, including extraneous variables
(this can be controlled with random assignment, and tested using pre-post test design)
what is pre-post test design, and what are the +/-?
pre test = check if the different groups of PPs perform a task equally before experimental manipulation
post test = check if experimental manipulation has changed performance
+ allows us to check if randomisation has worked
+ allows us to isolate the effect of the IV on the DV
- can alert PP to the thing we are interested in, and affect their behaviour
- can only test difference in performance, rather than absolute performance
solutions?
- disguise pre test with another task or context
- increase time between tests
+/- of within subjects design
+ no chance of differences in groups between conditions
+ control for individual differences
+ fewer PPs needed
- practice effects
- boredom
- carry-over effects
can have multiple IV levels
not restricted to only experimental/control group
can have multiple experimental groups
randomisation of multiple levels
within subjects = randomise the order that the PPs complete the different conditions, using latin square/counterbalancing to reduce order effects
between subjects = randomise condition PP completes
multiple IVs - factorial design
e.g., if measuring effect of coffee on memory as well as word length (long and short), match up each condition to each condition
decaf - group for long words, and group for short words
strong coffee - group for long words, group for short words
etc
what are mixed designs
participants complete some conditions
e.g., PPs complete only one caffeine condition, but see both long and short words
