Applied Research Methods Midterm

types of observational studies

cohort, case-control, cross-sectional, longitudinal

cohort study

used to compare the incidence of a disease between groups

example of cohort study

comparing rates of cancer (disease) between smokers and non-smokers (groups)

cohort

a group of people who meet a set of inclusion criteria e

example of cohort

free of cardiovascular disease and normal weight

______ numbers of participants to be enrolled

LARGE

case-control study

researchers select a disease (CASE) and compare 2 similar groups (case vs control group)

case group

group of participants WITH the disease

control group

group of participants WITHOUT the disease

example of case-control study

comparing a group of construction workers WITH cancer to a group of construction workers WITHOUT cancer

case-control study is used when…

the disease is RARE (ex. lou gehrig’s disease)

________ number of participants are enrolled in case-control study

small

cross-sectional study

conducted at a SINGLE point in time

example of cross-sectional study

a survey of behavioral risk factors among U.S. adults in 2021

in cross-sectional studies, __________ can cause bias or limit generality

non-response

longitudinal study

REPEATED observations of the SAME variable and the SAME participants over long periods of time

example of longitudinal study

observing body weight among obese children over a 10 year period

data collection in longitudinal studies…

occurs at MULTIPLE time points

types of experimental studies

randomized control trial, quasi-experiement

randomized controlled trial (RCT)

RANDOMLY assigns subjects into an experimental and control group with probability sampling

example of RCT

a clinical trial (ex. medical test)

example of control group

placebo group

in an RCT, participants have an _______ chance of being selected

equal =

RCTs use a _____ design

pre-post test design WITH a control group

example of pre-post design

recording weight PRE test and weight POST test

_____ can be established through an RCT

causality

causality

relationship between cause and an effect

criteria #1 for causality

cause MUST precede the effect in time

criteria #2 for causality

cause and effect MUST be correlated with each other

criteria #3 for causality

correlation between cause and effect CANNOT be explained by another variable

quasi-experimental study

LACKS the element of random assignment (probability sampling) to treatment or control

quasi-experimental study uses _______ sampling

NON-probability

non-probability sampling

participants DO NOT have an equal chance of being selected

quasi-experimental study is a _____ study

NON-randomized

quasi-experimental study ______ have a control group

does NOT

quasi-experimental study has ______ with experimental design

similarities

quasi-experimental study design

pre-post test design WITHOUT control group

quasi-experimental study uses only….

ONE intervention group

meta-analysis

a quantitative study design used to systematically assess previous studies to derive conclusions

steps in a meta-analysis

1. IDENTIFY a problem

2. REVIEW previous studies to determine inclusion or exclusion criteria

3. CALCULATE effect size

4. REPORT all the steps, outcomes, and conclusions in a meta-analysis

on what basis can studies be included/excluded in a meta-analysis

internal validity, sample size, study design, publication year, etc.

internal validity

eliminating all of the potential confounding variables (ex. third/extraneous variables) that correlate with both DV and IV

internal validity sets up…

strong controls to isolate other factors

third/confounding/extraneous variables

spurious and distorted associations between two variables

effect size

whether an experiment or trial has a statistically significant effect on a DV

example of effect size

cohen’s d= 0.8, the trial had LARGE effect on cholesterol levels

cohen’s d = 0.20

an experiment or trial yielded a SMALL effect on a DV

cohen’s d = 0.50

an experiment or trial yielded a MEDIUM effect on a DV

cohen’s d = 0.80

an experiment or trial yielded a LARGE effect on a DV

validity

how well a survey measures what its set out to measure

validity is the…

correctness, accuracy, and appropriateness of a measurement

criterion

a second test/other assessment procedure to examine the relationship between two variables 

criterion-related validity

quantified by the correlation coefficient

a HIGH correlation coefficient…

provides evidence for criterion-related validity

validity coefficient of 1.0…

allows researchers to perfectly predict the criterion

validity coefficient of -1.0…

shows a negative relationship

type of criterion-related validity

predictive validity

predictive validity

when measuring the relationship between two variables/predicting scores at a different time

example of predictive validity

measuring anxiety in high school vs. anxiety in college

construct validity

how well a test measures a construct that it was designed to measure

example of construct validity

if people who have depression show higher scores on Beck Depression Inventory than those without depression, this IS construct validity

reliability

consistency of a measurement and the correlation between items

cronbach’s alpha (alpha coefficient)

ranges from 0 to 1

the higher the cronbach’s alpha…

the MORE reliable the generated scale is

acceptable value of cronbach’s alpha is…

0.70

split-half procedure

usually odd vs. even items

split-half reliability coefficient is any value between…

0 and 1.0

desired split-half reliability coefficient

approaches 1

kuder-richardson approach (KR21)

KR21 value ranges in value from 0 to 1

the higher the KR21 coefficient…

the MORE reliable the generated scale is

population

an entire set of individuals, items, or data of interest 

example of population

all people with HIV

sample

a subset of individuals, items, or data of interest

population data are termed…

parameters

sample data are termed…

statisticse

example of sample

people with HIV in Chicago

most research is done on…

samples

descriptive statistics…

summarize sample results

inferential statistics…

generalize to populations

probability

a measure of the likeliness that an event will occur

probability formula

probability of an event happening = # of ways it can happen / total # of outcomes

two main types of selection of a subset of the population of interest

probability and non-probability sampling

probability sampling

each sample of the population has a KNOWN probability of being selected

non-probability sampling

each sample of the population is selected WITHOUT the use of probability

in probability sampling participants have…

an EQUAL chance of being selected

types of probability sampling

stratified, systematic

stratified sampling

researchers divide the entire population into different sub-groups based on characteristics like age, gender, etc.s

stratified sampling is used to…

examine a trend or issue across sub-groups

systematic sampling

researchers pick an interval number to select participants 

example of systematic sampling

population size = 100, sample size 10, interval number is 10

non-probability sampling

participants do NOT have an equal chance of being selected

types of non-probability sampling

convenience, purposive, snowball

purposive sampling

judgmental, selective or subjective sampling use to serve a very specific study need

convenience sampling

where subjects are selected because of their convenient accessibility and proximity

snowball sampling

used when the sample for the study is very rare or is limited to a very small subgroup of the population

in snowball sampling…

you can ask participants to nominate another person with the same trait/disease

research ethics

norms for conduct that distinguish between acceptable and unacceptable behavior

research ethics are also…

ethical standards and procedures for research or standards of conduct for scientific researchers

institutional review board (IRB)

an administrative body or an independent committee established to protect the rights and welfare of human research subjects

IRB committee is made up of…

3 to 5 committee members with research experience and expertise