Validity Flashcards
research strategy
general approach and goals of a research study, question/what is addressed determines which strategy is chosen
Research design
how the strategy is implemented: group vs individual, same/different individual, nb of variables
Research procedure
details on how the study will be done: how variables manipulated/ measured, nb of animals/ind. involved
descriptive research strategy
current state of a variable, no comparison between groups, relationship between variables is not examined
ex: how many neurons are in the hippocampus
Correlational research strategy
measures 2+ variables for every individual in the group,identify correlations between variables, measures relationships between variables(+,-, curve)
no manipulation
Non-experimental research strategy
examine relationship between variable, looking for difference between 2 groups, no ctrl group, one variable has 2 levels, no manipulation
Quasi-experimental research strategy
has ctrl group, groups cannot be randomly assigned, does not provide definitive answer about cause-effect,
Experimental research strategy
explain relationship between 2 variables, establishes cause and effect, has ctrl group, compares 2+ groups on dep. variable, randomly assigned group that are manipulated
internal validity
research that produces unambiguous explanation for the relationship between 2 variables, change in dep, v. must be due to change in independent variable
threat to internal validity
any factor that raises doubts or allow alternative explanations: cofound, obscuring variable
external validity
extent to which the experiemental result holds true outside the specific study
threat to external validity
any characteristic that limits the generalizability of the result: specific study/data, limited participants
Validity in experimental strategy
allows ctrl of the environment, conducted in ctrled and constant lab setting (high internal v) , exp research is often conducted in unfamiliar environments=> harder to generalize=> limits external v
validity in non experimental
high external validity: realistic environments
limited internal validity: environment is not manipulated or ctrld
generalizing across participants or subjects
(yes=threat to external validity) selection bias, convenience sample, volunteer bias, participant characteristics, cross-species generalization
generalizing across study features
(yes=threat to external validity) novelty effects: influence from being in novel environment
multiple treatment interference,
experimenter characteristics
generalizing across different measures
(yes=threat to external validity)
sensitization : monitoring a behaviour can alter behavior
timing of measurment
cofounding variables
threat to internal validity
change with the ind. v = changes in the dep. variable , difficult to remove
ex: assignment bias, environmental/time related variables, observer bias
Obscuring variables
threat to internal validity,
makes changes in dependent variable hard to observe, lead to measuremnet error, noisy data
ex: ineffective manipulation, measurement error, excessive variation in data
criteria for generalization
- controlled comparison rule
- sampling rule
- operational definitions rule
controlled comparison rule
comparing data between experimental and ctrl groups
when comparison is ctrld => internal validity
sampling rule
adequate sample size to control for random sampling error, unbiased sample is representative or pop., can generalize broad principles
operational definitions
determines what principles can be generalized, replicated experiments can be used with different operational definition to generalize
