1/47
Looks like no tags are added yet.
experiment
involves the manipulation of one or more factors and measurement of the effects of this manipulation on behavior
internal validity
external validity
reliability
sensitivity
independent variables
the variable that is being manipulated
dependent variables
the variable(s) being measured
confound variables
variables that vary completely with the independent variable such that we can’t tell which produces the effect
control variables
something you keep constant across the groups (so that it can’t affect results)
between-subjects design (independent groups)
has different people in your different groups
why choose a between-subjects design?
random selection
random assignment
have a convenient sample (take whoever showed up) then randomly place them into groups
matched groups design
used when there is a variable that you want to make sure doesn’t influence results but you can’t keep it constant
pretest
post-test
experimental condition
receives the independent variable
control condition
does not receive the independent variable, they’re there for comparison purposes
demand characteristics
experimenter effects
subject loss in an experiment
regression to the mean
overtime groups scores are going to drift toward the mean
placebo
double-blind experiments
neither the participant or researcher should know which condition they are receiving
nuisance factors
something other than the independent variable that could influence the results (but we don’t want it to)
within-subjects design
the same people are exposed to different levels of the independent variable
advantages to within-subjects design
use fewer subjects, short/long-term, hypothesis driven, cheaper, statistical sensitivity, independent effects
use fewer subjects
rare with enough for a group but not a massive study, matching groups
short/long-term
good for something that is short or long period of time
hypothesis driven
what you are studying will determine if within-Ss work
cheaper
because of needing fewer subjects: money in case you are paying participants/equipment and training
statistical sensitivity
how likely you are to find a difference when there is one
independent effects
independent variable may have a different effect when studied within-Ss
disadvantages to within-subjects design
practice effects, fatigue effects, order effects, anticipation effects
practice effects
the more you redo a study, if results are too similar/really strong, may need to do between-Ss
fatigue effects
doing something back to back, you get tired
order effects
presentation of the conditions affects performance
anticipation effects
knowing what is coming next for a study, so it’s not a true response and the results aren’t true
complete within-subjects design
the effects are balanced across individuals
incomplete within-subjects design
the effects are not balanced out across individuals but are balanced out across groups
block randomization
how many levels do you have then that will be multiple of the blocks (sets of trials), of each block they will be a multiple, and each block will be randomized
ABBA counterbalancing
Used for something that will take a long time. Often used in therapy. You get A then B then B then A, presented over time
complete counterbalancing
Groups of people getting the all possible different orders rather than individual
latin squares
Used if you have 4 or more conditions, you will have a square (four rows, four columns)
Have a reference line to look at but not part of the square, label the conditions (ABCD -> 1,2,3,N)
First line is always 1, 2, N, 3 -> ABDC
Do math to create the rest of the lines, +1. BCAD, CDBA, DACB
random starting order with rotation
complex designs
when you have two or more one-way variables
complete factoral design
occurs when each level of one independent variable is combined with each level of a second independent variable
incomplete factoral design