Research Methods Exam 2

population

the entire group the researcher wants to generalize to

sample

the group actually studied

census

measuring every member of a population

external validity

how well findings generalize beyond the sample

representative sample

a sample that resembles the population on important characteristics

convenience sampling

choosing people who are easy to reach

self-selection sampling

people volunteer themselves

probability sampling

every member of the population has a known, nonzero chance of selection

simple random sampling

every member has an equal chance of being selected through a random process

cross-sectional correlations

two variables measured at the same time

autocorrelations

the same variable measured across time

cross-lag correlations

one variable at Time 1 predicting another variable at Time 2, and vice versa

mediators

why or how two variables are related

moderators

explains when, for whom, or under what conditions the relationship occurs

independent variable

the variable the researcher manipulates

dependent variable

the outcome that is measured

levels of the IV

the different conditions created by the researcher

why is probability sampling the gold standard?

provides strong representativeness and external validity by giving all members of the population a known chance of selection

what are the main drawbacks of convenience and self-selection sampling

can produce systemic bias and reduce the validity of results

does a large sample size fix a biased sample

no due to poor sampling methods

what does increasing sample size improve?

improves precision, has smaller margin of error

what improves accuracy and representativeness in a study?

better sampling methods, not just larger samples

what does margin of error tell us about a study?

reflects precision, not if the sample is unbiased or representative

which sampling method is most representative?

probability sampling

why is convenience sampling a problem?

risk of bias, not representative of the population

why does sample size not fix bias?

reduces random error, not systematic error

what does margin of error tell us?

reflects precision, not whether the sample is unbiased

why is external validity especially important for frequency claims?

because they aim to estimate how common something is in a population

what type of variables are required to compute a Pearson correlation ( r ) ?

both must be quantitative

is a standard Pearson correlation appropriate if one variable is categorical?

no, it is not appropriate when one variable is categorical

why is it important to know whether a variable is quantitative or categorical?

it determines which statistical analysis and graph are appropriate

correlation

a relationship between two measured variables

positive correlation

as one variable increases, the other also increases

negative correlation

as one variable increases, the other decreases

strength of correlation

absolute value of r, not whether + or -

what can scatterplots reveal?

direction, strength, outliers, curvilinear patterns

why would a curvilinear relationship have an r value near zero?

r only captures linear relationships

causal critera

covariance, temporal precedence, internal validity

correlation does not equal:

causation

possible explanations for a correlation:

A causes B, B causes A, third variable causes both A and B

why correlational studies cannot establish causation:

do not manipulate variables, do not use random assignment, can’t establish temporal precedence, can’t rule out third variables

moderator

explains when, for who or under what conditions the relationship occurs

multivariate design

includes more than two measured variables

multiple regression

include several predictors at once, statistically control for possible third variables, estimate the unique contribution of one predictor while holding others constant

mediators

why or how two variables are related

random assignment

each participant has an equal chance of being places in any condition

random assignment:

helps create equivalent groups, supports internal validity, not the same as random sampling