Research control Flashcards
constancy of conditions
try your best to control
random assignment
everybody in population, has an equal chance of randomly being picked to be place in either group 1 or 2
subjects exposed to more than 1 level treatment
- repeated measures, same group got to have all the different treatments over and over, serve as their own controls
- extraneous variable are going to same
homogeneous sample
you know what youre extraneous variables make sure each person have same characteristics to have all variables you identified
build in extraneous variables into
controlled and intervention group
randomized block design
- ruling out some of impact of extraneous variables, of how you can collect data
- researcher not doing anything
matching
make sure both groups have same characteristics
analysis of covariance
test the main and interaction effects of categorical variables on a continuous dependent variable, controlling for the effects of selected other continuous variables, which co-vary with the dependent
methods of evaluation of adquacy of research
- internal validity
- extenral validity
internal validity
attained in a study when findings can be shown to result only from effect of independent variable of interest and cannot be interpreted as reflecting effects of extraneous variables
if internal validity is strong you can say
_____is why_____
external validity
results can confidently be generalized to situations outside specific research setting
history
occurrence of extternal events that can take place concurrently with independent variable
selection
biases that result from preexistence difference between group and if they aren’t randomly assigned going to have non equal groups and defer subtly that you are not aware of
may cause problem
maturation
- when you are looking at people who are changing over time
ex: could have older folks changing declining
ex: young adult going through relationship and maturing
testing
they might perform better
don’t know if independent variable effected it
instrumentation
-if researcher uses different measurements in beginning and later in study
-if you have administrators (way administer test) differences in instrumentation
difference to data and validity
mortality
dropped out of study, expecting so many people to look at outcome and people leave you can’t do stats or can’t do it as well
threats to internal validity
- history
- selection
- maturation
- testin
- instrumentation
- morality
threats to external validity
- Hawthorne effect
- novelty effects
- interaction of history and treatment effects
- experimenter effects
- measurement effects
opt always for
stronger internal validity
Hawthorne effect
occurs when subjects change their behavior because they know they are being observed
novelty effects
- results of a study are due to the novelty (or newness) of a treatment
- pose threat to external validity because they make it difficult to know if the results of the study are due to a treatment that works or due to the novelty of a treatment.
experimenter effects
influence that scientists who conduct an experiment have on performance of participants and the interpretation of results.
measurement effects
merely measuring or questioning an individual’s intentions or anticipated regret changes their subsequent behavior