a. One way to check reliability is test-retest. Run the same test again
and again and get the same results. Can have issue if the unit is in
b. Alternative form method: use multiple alternative measures to see if
get same score
1. Operationalization: turning your concepts and ideas into quantities is very
a. Needs to be clear, rigorous, and to justify your choices.
2. Need to explain causal relationship in hypothesis
3. Measurement is the systematic observation and re
1. Non-experimental designs
i. High external validity
ii. Can examine large number of cases at the same time.
iii. More likely to detect general patterns
i. Very difficult to control and randomization is a problem.
1. Disadvantages of case studies
a. Very small sample makes it harder to generalize. Questionable
b. Make sure select unbias sample. Cant just choose the best case for
c. Replicability: hard to replicate because takes a lo
1. Research Design for next week
a. Need to first come up with hypothesis
b. Figure out way to test the hypothesis
2. What are the characteristics of a good hypothesis?
a. Causal nature
b. General: explains a number of different cases
c. Empirical (testab
1. Time-series research designs
a. Nature creates a quasi-experiment by imposing a treatment on a
population. Generally takes a pre-test post-test design.
b. Cant hold all variables constant, and intervention may not be
randomly timed or targeted. Usually
1. Inductive reasoning takes the data and creates a theory. Deductive
reasoning, working without data, to create a theory through premises.
2. Spurious correlation: when two things are correlated with not because of
3. Research design is roadma
1. What is an experiment? An experiment is an artificial situation in which
you look for a relationship.
a. Experiment in humans is when you subject a group of people to
treatment or not.
2. Characteristics of an experiment
a. Get a representative sample
a. Intervening variable: a variable caused by the IV that youre
studying, which in turn causes the DV.
i. Alternate: a variable that changes the effect of the IV that
b. Control Varibale: something that probably causes the DV but that
i. Trying to create probabilistic models expect variation in
b. Scientific models oversimplify the truth of any social phenomenon.
i. Again, looking for a general pattern of behavior. Looking to
find three or factor that affect the behavior of most
1. Science is one way, and historically the most successful way, that
knowledge about the physical world and the things in it are obtained.
a. Scientific knowledge is created by scientific process. It is that
knowledge that survives the process.