Research Final

Population

all possible elements of population

Parameter

characteristics of the population

Sample

subset of the population

Statistic

characteristic of the sample

Validity threats related to sampling

noncomparable samples, history effects, attrition, regression to the mean

Sampling Variability/Sampling Error

chance variations in observations from one sample to another

reflects difference between sample statistics and population parameters (characteristics of the sample vs the population)

ex: getting different averages from sampling 1000 people and then another 1000 people, the only difference is chance

Statistics allow us to….

estimate and contend with uncertainty created by chance variations, most important idea

allow us to deal with the uncertainty

sampling frame

concrete listing of the elements in the population

allows for probability sampling

sampling unit/element

individual component of measurement

probability sampling

procedure where the likelihood of selecting any one person in the population can be determined

everyone should have the same chance of being picked

simple random sampling

probability sampling

unsystematic random selection of individual from the sampling frame

systematic random sampling

probability sampling

systematically selecting individuals from randomly sequenced sampling frame (picking every 10th)

stratified random sampling

probability sampling

randomly sampling within defined sampling frame subgroups, random sample proportionate to the size of the groups

cluster sampling

probability sampling

randomly sampling groups within one sampling frame and then randomly sampling from within those groups

nonprobability sampling

people have different likelihoods in being included

accidental sampling

nonprobability sampling

selection based on ease of researcher

convenience or availability sampling

purposive sampling

nonprobability sampling

selection designed to obtain a sample that has certain characteristics, intentionally recruiting on some characteristic

quota sampling

nonprobability sampling

sampling based on categories in the population

exactly like stratified random sampling, but without access to the sampling frame

snowball sampling

nonprobability sampling

initial participants asked to recruit additional participants that meet the qualifying criteria

often used with demographics who are difficult to recruit on your own

can introduce homogeneity bias, people often recruit their friends

structure oriented

cultural consideration

examine similarities and differences in relationships between variables across cultures (what does a culture “look-like”?)

level oriented

cultural consideration

different in magnitude of variables across cultures (what does the variable of interest look like?)

attitudes, norms, etc.

decentered

cultural consideration

culturally diverse perspective taken at the start of the study

consult with researchers from cultural background

convergence

study designed in order to remove cultural influences of existing related research

results later compared, patterns illustrate potential bias

can get a sense of the underlying trend independent of culture

cultural factors not taken into account initially

conceptual equivalence

the aspects of the test/assessment mean the same thing

ex: BDI vs. PHQ9

scaler equivalence

the words in one language mean the same thing in another

back translation- method to identify bias, translating language back and forth blindly to id errors

Chicago tribune fun fact

newspaper said Thomas Dewey won the election, but had only polled higher SES people who had a phone

Harry Truman actually won the election