research methods exam 1

acronym for purpose of scientific research

DUH

what does DUH stand for

D- discovery (interventions, treatments, practices, drugs)

U- understanding (theory developments)

H- hone in on evidence based practices (examine and compare)

quantitative definition

hard, reliable data, objective, researchers POV

qualitative definition

rich, deep meaning, subjective, participants POV

external validity is?

sampling techniques that capture the population/representative of sample

internal validity is?

how well the experiment is designed so that it rules out alternate explanations of the results

construct validity is?

reliable of the instruments or measurements

statistical validity is?

are we using the appropriate statistical test/ are we analyzing the stats correctly

what are we most interested in?

IV

active IV

the researcher assigns to subjects

attribute IV

existing characteristics of subjects and are not assigned by the researcher

what is the outcome that measures the IV

DV

the extraneous variable is?

any variable other than the IV and DV that you are not investigating but can potentially affect the DV of the research study

what is the confounding variable?

type of EV that relates to the DV and IV in a way that distorts or confuses the relationship

three types of research in medicine and healthcare

correlational, qualitative, descriptive

meta analysis

strongest design, synthesis of the literature in which a researcher uses statistical techniques that summarizes the results of several studies in a single weighted estimate, more weight given to results of studies with more events and higher quality

systematic review

review of literature to address a defined question in which a researcher uses a specific and methodical approach to identify and summarize studies

true experimental

random assignments/ controlled trials

best control for EV

researcher examines the cause and effect relationships between IV and DV by comparing two of more randomly assigned groups

quasi experimental

no active IV

no random assignment

researcher examines the cause and effect relationships between IV and DV by comparing two or more non randomly assigned groups

cohort

observational

attribute

opposite of case control

observational study where the researcher tracks a group who are exposed to a suspected risk factor and group who was not, to ascertain whether the exposure is likely to cause a specific disease or outcome

case control

observational

attribute

looks backwards

less effective at determining causality because of external influences

where a researcher compares a group of people with a disease or have experience with an adverse outcome with a group of people without the disease or who have not experiences the same adverse outcomes to see if the two groups have differing exposures that account for the differences

cross sectional

observational attribute

describes characteristics of a population at a particular point in time or over a short period

compares two or more groups from that population across some variable

case report

a descriptive non comparative study that analyzes a group of people who have a disease (rare) or who have been exposed to a risk factor or intervention

observational

attribute

weak casual evidence

animal lab studies

weakest

done the most

does not translate to humans

helps determine the plausibility of clinical studies with people

which type of research is the most robust

studies of studies

seek out first

meta and systematic

what are the experimental designs

true and quasi

what are the observational designs

cohort, case, cross sectional

what is the non human designs

animal lab

if IV is assigned by the researcher the IV is? the study is?

active and experimental

if the IV is not assigned by the researcher the IV is? the study is?

attribute and observational

if the subjects of a experimental and active IV is randomly assigned it is

true experimental

if the subjects of a experimental and active IV is not randomly assigned it is

quasi experimental

if the variables are exposure then outcome it is a

cohort

if the variables are outcome then exposure it is a

case control

if the variables are exposure and outcome at the same time it is a

cross sectional

therapy

determines the effectiveness of a treatment or interventions for a given condition, disease, or disability

most appropriate research design for therapy is

true experimental

etiology

asks about the causes or risk factors associated with the disease or conditions

most appropriate research design for etiology is

cohort

diagnosis

determines the accuracy of a diagnostic tool or test

most appropriate research design for diagnosis is

cohort and cross sectional

prognosis

examines the probable outcome and progression of a disease or condition

most appropriate research design for prognosis is

cohort

what is PICO

problem or patient population

intervention

comparison or control

outcome

population validity

ecologically validity

accessible sample/sampling frame

who we could potentially sample

selected sample

who we sample

actual sample

who participants

simple random

everyone in the population has an equal chance of being selected for the sample
the list has to be random

systematic random

draws a random sample from the target population by selecting units at regular intervals starting from a random point

stratified with equal proportions

randomly selecting a sample from within certain strata, or subgroups within the population. subgroup is separated from the others based on a common characteristic

different number of people in each group

stratified random with different proportions

randomly selecting a sample from within certain strata or subgroups within the population

same amount in each group

cluster

natural occurring clusters are formed and then the researcher picks a single cluster

convenience

simply recruiting anyone you have access to

quota

separate people into strata or groups and then enroll anyone willing to participate in the study

groups then you get who you get

purposeful

people who are considered experts on the topic of interest are identified and recruited for the study

target the experts then you get who you get

qualitative

snowball

iterative process whereby the researcher establishes trust and rapport with person who have knowledge about the topic

target population is hard to get to/does not want to be found

created by word of mouth

non probabilistic sampling techniques

convenience, quota, purposeful, snowball

probabilistic sampling techniques

simple, systematic, stratified, cluster

R

random assignment

NR

not randomized

O

observation

X

intervention or treatment

~x

no intervention or the usual treatment

E

experimental or intervention group

C

control

M

matching

quasi one group post test only design

NR, E, X, O1

quasi One group pretest-posttest design

NR, E:, O1, X, O2

quasi Posttest Only Nonequivalent groups design preferred over one group pretest posttest

NR, E:, X, O1

NR, C:, ~X, O1

quasi Pretest-Posttest Nonequivalent Comparison group designs

NR, E:, O1, X,O2

NR, C:, O1, ~X, O2

better quasi Single group time series design  (with temporary treatment)

NR, E:, O1O2O3O4 X O5O6O7O8

better quasi Single group time series design ( with continuous treatment)

NR, E:, O1O2O3O4(pre) X O5 X O6 X O7 X O8 (treatment)

better quasi Multiple group time series designs (with temporary treatment) 2 groups

NR, E:, O1O2O3O4 X O5O6O7O8

NR, C:, O1O2O3O4 ~X O5O6O7O8

true Posttest only control group design

R, E:, X, O1

R, C: ~X, O1

true Pretest-posttest control group design most thought of

R, E:, O1, X, O2

R, C:, O1, ~X, O2

true Solomon four-group design very strong

R, E1:, O1, X, O2

R, E2:, X, O2

R, C1, O, ~X, O2

R, C2, ~X, O2

true Crossover study- compared to groups and themselves/ less people

R, E:. O1, X, O2 CROSSOVER, ~X, O3

R, C:, O1, ~X, O2, CROSSOVER, X, O3

true matching Posttest only control group design

MR, E:, X, O1

MR, C: ~X, O1

attrition bias

When there are systematic differences between the trial groups because participants have dropped out or have been excluded during follow up

counter measure for attrition

random assignment, blinding, quality assurance

allocation bias

When there is a systematic difference in how participants are assigned to the intervention and comparison group in a clinical trial

countermeasure for allocation

random assignment

testing

When there is fatigue or a possible carryover effect from the pre-test to the post-test

counter measure of testing

control group and timing

maturation

When participants naturally change as a function of time rather than due to the IV

Young and old susceptible to this

Age wealth health education

countermeasure for maturation

control group

performance bias

When members of one group change their behavior when they realize that they are in a certain group, or when there are differences in care provided to the groups because those administering the treatment have certain expectations

Modify behavior with control/treatment

Researcher treats patients differently

counter measure for performance bias

blinding quality assurance

history

When something external happens right before the study or between the pre-test and post-test that alters how one responds to the assessment

More subtle and unique

counter measure for history

control group

detection bias

researchers have preconceived notions of the treatment

fail to follow protocol/people who do not follow instructions

characteristic of a subject makes it difficult to measure

countermeasure for detection bias

blinding quality assurance

phase 1

testing for safety

20-80 people

low dose

phase 2

seeing if actually works

optimal dose

side effects

100-300

phase 3

does it work better/compared to standard treatment

double blind single blind crossover

longer

placeboes

300 with condition

apply to FDA

phase 4

what else?

monitor

1000+

cost effectiveness

trends

prospective cohort study

start with exposure and then the researcher comes in and follows the group over time to see if DV occurs

retrospective cohort study

exposure, DV occurrence?, researcher comes in