Please enable JavaScript.
Coggle requires JavaScript to display documents.
Psych - Research Methods, Experimental Design, Descriptive Statistics,…
-
Experimental Design
Repeated measures
-
individual differences, half as many PPs required
order effects - the PPs do more than 1 condition, affects how they perform in the next condition, e.g. tired, skilled, etc.
conuterbalancing - one group of PPs does condition A, then condition B. Other group does condition B, then condition A (ABBA)
independent measures
-
order effects, demand characteristics
-
matched pairs
-
order effects, individual differences
time consuming, expensive, not all individual differences addressed
-
Aims + Hypothesis
Hypothesis
theories tested by research, testable
-
-
-
directional hypothesis
predicts change, not how it will occur
-
-
Types of Experiments
experiment
controlled situation, researcher manipulates 1 variable, observes change on another variable, keep other variables constant
laboratory experiment
controlled setting, researcher manipulates IV, discover effects on DV
control, internal validity, replicatability
external validity, demand characteristics, mundane realism
field experiment
everyday setting, IV manipulated by researcher, observe effects on DV
external validity, mundane realism
control, internal validity, replicapability, informed consent (ethics)
natural experiment
IV changes, not manipulated by researcher, observe effects on DV
confounding variables, internal validity (cause/ effect)
external validity, ethical
quasi experiment
IV not controlled or changed by researcher, effects on DV observed
-
no random allocation, causes confounding variables
variables
-
dependent variable (DV)
changes, observed by researches
-
Sampling
sampling techniques
random sampling
-
time consuming, difficult to conduct, bias possible in small amount of PPs
-
systematic sampling
every Nth PP is selected, a sampling frame is produced
-
-
time consuming, still small chance of bias
stratified sampling
researcher divides people into groups representative of general population. E.g. in a class of 40 with 30 male and 10 female students, a stratified experiment with 20 of them would have 15 men and 5 women.
researcher bias, representative
cannot reflect all the population, not fully representative
opportunity sampling
-
-
representative, researcher bias
-
sample
-
-
representative, generalisable
-
-
-
-
-