Quantitative Research Design for Nursing page 3

porkchop205's version from 2016-10-24 15:43

Section 1

Question Answer
blinding or maskingprevents expectation bias from awareness. conceals infromation from all those involved in the study.
performance biassystematic differences in care provided to members of different groups apart from any intervention
detection or discernment biassystematic differences between groups in how outcomes are measured, verified or recorded.
open studynot blinded
closed studyinvolves masking or blinding
single-blind studyonly one group of people (eg study participants) are blinded
double-blind studytwo groups are blinded (eg both those providing the intervention and those who are receiving it)
blinding vs maskingthe term blinding may be offensive. it may also be confusing to the reader if the population is sigh-impaired.
main effectseffects from experimentally manipulated variables
intervention effectseffects from combining treatments

Section 2

Question Answer
post-test only designdata on dependent variable are collected only once
pre-test post-test designmeasures for basline before intervention and again after intervention
repeated measures designanother name for pre-test post-test design
after-only designanother name for post-test only design involving collection of data points at multiple post-intervention points
crossover designparticipant serves as their own control
factorial designmanipulation of two or more variable simultaneously

Section 3

Question Answer
basic post-test only designuseful when outcome is not relevent until after the intervention is complete
basic post-test design drawbackdoes not permit evaluation of whether or not the two groups were comparable at the outset
basic pre-test post-test designused when focus of intervention is on change
basic pre-test post-test designused when researcher wants to assess both group differences (experimental comparison) and change within groups (quasi-experimental)
basic pre-test post-test design drawbackPretest itself can affect outcomes
Multiple intervention designcan be used to disentangle effects of different components of a complex intervention, or to test competing interventions
Multiple intervention design drawbacksrequires larger sample size than basic designs
Multiple intervention design drawbacksMay be at risk to threats to statistical conslusion valididity if A and B are not very different (small effects)
Wait-list designattractive when there is patient preference for the innovative treatment
Wait-list designcan strengthen inferences by virtue of replication aspect for second group
Wait-list design drawbackscontrols may drop out before getting deferred treatment
Wait-list design drawbacksnot suitable if key outcomes are measured long after treatment or if there is an interest in examining long-term effects
Crossover design appropriate if there is no expectation of carryover effects from one period to the next (effects should have rapid onset, short half-life
Crossover design useful when recruitment is difficult - smaller sample is needed; excellent for controlling confounding variables
Crossover design drawbacksoften cannot be assumed there are no carryover effects
Crossover design drawbacksif the first treatment "fixes" a problem for participants, they mae not remain in the study for a second one
Crossover design drawbackshistory threat to internal validity a possibility
Factorial designefficient for testing two interventions simultaneously
Factorial designcan be useful in illuminating interacton effects; most useful when strong synergistic effects are expected
Factorial design drawbackspower needed to detect interactions could require larger sample size than when testing each intervention separatelyra

Section 4

Question Answer
randomization in crossover designparticipants must be randomly assigned to different orderings (group a gets treatment 1 then 2; group b gets treament 2 then 1)
washout period a period of no treatment exposure for a certain period of time is used (to remove the possibility of carryover effects)
Hawthorne effectplacebo-type effect caused by people's expectations simply from being included in the study (people change behavior because they know they are being watched/studied)
nonequivalent control group pre-test post-testused in quasi-experimental design - two groups whose intervention outcomes are measured before and after intervention
comparison group in quasi-experimental, this is term often describes the control group
propensity scorecaptures the conditional probability of exposure to a trement given various pre-treatment characteristics
historical comparison groupcomparison data are gathered before implementing intervention
time series designinformation on the dependent variable is collected over a period of time before and after intervention; oftne used in single-subject experiments
quasi-experimental research design typesdose-response analysis; nonequivalent pretest-post test design; PRPP; one group pretest-post test design; time series designs, dose-response design

Section 5

Question Answer
Nonequivalent control group pre-test post-test designuse when an entire unit must get the intervention and a similar unit NOT getting the intervention is available
Nonequivalent control group post-test only designreasonable choice only when there is some a priori knowledge about comparability of groups with regard to key outcomes
One-group pretest-posttest designresonalble choice only when intervention impact is expected to be dramatic and other potential causes have little credibility
Time series designgood option when there are abundanat data on key outcome in existing records
Time series designaddresses maturation threat and change from secular trends and random fluctuations
Time series noneqivalent control group designattractive when entire unit/institution adopts the intervention and a similar unit is not, and if comparable data are readily available in records of both
Time series with withdrawn and reinstated treatmentatttractive if effects of an intervention are short-term

Section 6

Question Answer
Nonequivalent control group pre-test post-test drawbackshistory threat possible
Nonequivalent control group pre-test post-test drawbacksselection threat remains a nearly intractable problem, but no less so than with no pretest
One-group pretest-posttest design drawbacksextremely vulnerable to history threat
Time series design drawbackscomplex statistical analysis that is most appropirate with very large number of data points (100+)
Time series design drawbackshistory threat remains and (sometimes) selection threat if the population changes over time
Time series noneqivalent control group design drawbacksselection threat as two units or institutions are rarely identical
Time series noneqivalent control group design drawbacksanalyses may be complex
Time series with withdrawn and reinstated treatment drawbacksmay be untenable to assume there are no carryover effects
Time series with withdrawn and reinstated treatment drawbacksmay be difficult ethically to withdraw treatment if efficacious

Section 7

Question Answer
statistical process controlused to assess effects when data is collected sequentially over a period of time before and after implementation of an intervention or practice change
AB designA= baseline data gathering phase; B=intervention phase
ABA designwhen AB is done, then treatment is withdrawn
ABAB designwhen ABA is done, then treatment is reinstated
Correlationdoes not prove causation

Section 8

Question Answer
Nonexperimental/Observational Researchcorrelational cause-probing research; descriptive research
Nonexperimental/Observational Research--Correlational cause-probing researchRetrospective designs; Prospective nonexperimental designs; Natural experiments; Path analytical studies
Nonexperimental/Observational Research--Descriptive ResearchDescriptive Correlational Studies; Univariate Descriptive Studies
Retrospective designspresent phenomenon linked to the past. Researcher begins with dependent variable and examines whether it is correlated to one or more previously occuring independent variables
Prospective nonexperimental designs
Natural experiments
Path analytical studies
Descriptive Correlational Studies
Univariate Descriptive Studies


Qualitative Research Design