U2: Foundations: Methods and Approaches Flashcards
experiment
investigation trying to understand relations of cause and effect
independent variable
manipulated variable
dependent variable
measured variable
control variable
variable that is constant
representativeness
degree which sample accurately reflects the population
sampling bias
bias that occurs when choosing sample
types: bias of selection, self-selection bias, prescreening/advertising bias, healthy user bias
bias of selection
unrepresentative selection, occurs when people are selected in a physical space
self-selection bias
when subjects have some control over whether they participate
pre-screening/advertising bias
how volunteers are screen/when advertising is placed skewing the sample (i.e. “stop smoking treatment” add, leads to you getting people who are motivated even without the treatment)
healthy user bias
when study populations is in better shape than the general population
single-blind design
subjects don’t know if they’re in experimental or control group
double-blind design
neither subjects nor researchers know if the latter is in the experimental or control group
placebo
sugar pill, used to trick control group to thinking they’re getting the treatment
correlation research
assessing the degree of association between two or more variables
confounding variable/third variable/extraneous variable
unknown factor that plays a role
longitudinal study
correlation study, study over a long period of time with the same subjects
cross-sectional study
correlational study, testing a wide variety of subjects from different backgrounds to increase generalizability
case studies
clinical research, study of single individual to allow for general conclusions about other cases
survey
used for correlation research, way to accumulate data through questionares/interviews
conceptual definition
theory or issue being studied
operational definition
the way a theory/issue will be directly observed/measured in the study, has to be internally and externally valid
internal validity
certainty with which the result of an experiment can be attributed to manipulation of the independent variable and not other extraneous variables
external validity
extent to which findings can be generalized to other contexts
naturalistic observation
observation of subjects in natural setting, outside of a lab
descriptive statistics
summarize data
inferential statistics
testing hypotheses about data and determining how confident to be about inferences about the data
central tendency
descriptive statistics, characterizes the typical value in a set of data
positive skew
most values are on lower end, some are larger values
negative skew
most values on higher end, some are lower values
standard deviation
average dispersion of numbers around the mean
in typical distribution, 68% of all scores are within one standard deviation and 95% are within two standard deviations
pearson correlation coefficient
describes the linear relationship between two attributed scale of -1, 0, 1 1 = perfect positive correlation -1 = perfect negative correlation 0 = attributes are not related
positive correlation
as X increases, Y increases
negative correlation
as X increases, Y decreases
sample size
N or n
number of observations or individuals measured
the larger the sample size the more accurate it will be to the general population
null hypothesis
a treatment had no effect in an experiment
inferential statistics will allow possibility of rejecting null hypothesis with known level of confidence
alternative hypothesis
a treatment had an effect
alpha
the accepted probability that a result of an experiment can be attributed to chance rather than manipulation of the independent variable
this is always possible so alpha is usually set to 0.05 – which means that experiment’s results are statistically significant if probability of the results happening by chance in less than 5%
type I error
false positive, conclusion that a difference exists when it actually doesn’t
type II error
false negative, conclusion that there is no difference when there actually is a difference
p-value
probability of making a type I error, indicates that results are statistically significant (not only due to chance)
deception
used only if informing participants of nature of experiment might bias results
stanley milgram
obedience experiment, convinced people they were administering painful electric shock to other participants (who knew the true nature of the experiment)
confederates
people who pretend to be participants but actually know the nature of the study
institutional review boards (IRBs)
assess research plans before approval to make sure it meets ethical standards
ethical standards
- informed consent: participants must give consent and can leave the study at any time
- debriefing: after study is done. participants must be told the exact purpose of their participation
- confidentiality