PSY 230 Exam 2 Review

Biased (nonprobability sampling)

unrepresentative sample of the population

Unbiased (probability sampling)

representative sample of the population

Convenience Sampling

selecting participants because their easily accessible

Biased sampling techniques

convenience sampling, self-selection, non-response bias, purposive sampling, snowball sampling, quota sampling

Unbiased sampling techniques

Simple random sampling, systematic sampling, cluster sampling, multistage sampling, stratified random sampling, oversampling

Self-selection sampling

participants volunteer their participation/responses

Non-response Bias

participants that didn’t participate in a study differ from those who did

Purposive Sampling

researchers look for a specific population, but don’t get participants randomly

Snowball Sampling

researchers recruit by asking participants to refer others

Quota Sampling

researchers choose participants to meet a certain number

Simple random sampling

randomly pick people to be a part of the sample group

Cluster sampling

sample from naturally divided groups, then randomly select whole clusters

Multistage sampling

randomly select a cluster, then randomly select participants in the selected clusters

Stratified random sampling

divide population into groups, then randomly select an equal number of participants from each group

Systematic sampling

select every kth person to be in your sample

oversampling

recruit more than an equal number of participants to overrepresent a group in a sample

nonrandom samples =

lower external validity

sample representativeness matters more than

sample size

larger sample size doesn’t =

higher external v. (unless the sample is truly representative)

Open ended question PRO’s

research can focus on answer themes, participants can respond however they want

Open ended question CON’s

time-consuming to record, responses may be difficult to interpret

Forced Choice Question PRO’s

refined responses

Forced Choice Question CON’s

options may not be representative enough

Contingency Questions

questions only opened if you selected a specific response above

Likert Scale

used to indicate agreement (MUST HAVE 7 OPTIONS)

Likert Types

Scales that don’t follow the Likert scale exactly (less than 7 choices, poor→ excellent choices)

Double-barreled question

1 question that asks about two different things

negatively worded question

question with contradicting words in them

Response set

question answers are the same for multiple questions

Acquiescence

specific type of response set (participant may respond positively to every question to please researcher)

fence-sitting

participant answers are neutral/indecisive

social desirablity

desire to present yourself in a positive way

Reverse coded items

ask questions that contradict another question to see if participants are paying attention

Reactivity

participant behavior changes when they know they’re being watched

Observer bias

researcher’s expectations unconsciously influence their interpretation of participant outcomes

Operationalizing

defining how a variable will be measured in a study AND deciding the appropriate measures

Self-report Measure

answer about yourself is the outcome measure

Observed/Behavioral measure

examine an outcome by recording/observing actions/behaviors

Physiological Measure

biological data

Nominal Measurement Scale

labels/categories of different groups (Asian, African American, Caucasian)

Ordinal Measurement Scale

categories arranged in order/rank (large, medium, small)

Interval Measurement Scale

rank order of measurements using equal intervals to show differences (40 degrees, 50 degrees, 60 degrees)

Ratio Measurement Scale

rank order of measurement using equal intervals with a true zero point (0 is symbolic, the least amount of classes you can miss is 0)

Categorical Variables (qualitative)

describe a quality of certain things/people

Quantitative Variables

describes a quantity of something

reliability

how consistent a measure’s results are

Test-retest reliability

how consistent scores are on a particular measure over time (test and get same results over time)

Interrater Reliability

how well raters agree with one another

Correlation Coefficient

tells you the direction and strength of the relationship between 2 variables (-1 to +1)

Cronbach’s Alpha

measure internal reliability, tells you if test questions are measuring the same topic (ex: anxiety), r=.7 is the absolute minimum of reliability accepted

Internal Reliability

measure of consistency between items on the same scale

Beneficence and non-maleficence

research benefits society and no harm/suffering during research

fidelity and responsibility

professional behavior in the work field and establish trust with people you work with

integrity

be accurate, truthful, and honest with people

justice

all groups are treated fairly

respect for persons

participants CAN consent, maintain confidentiality/privacy, and NO coercion