Quantitative Research Flashcards
What is a paradigm?
patterns of beliefs and general assumptions
Name some alternative terms for a paradigm
research traditions
world views
methodologies
name some quantitative methodologies used in nursing research
quantitative - positivism / post positivism
name 4 strengths of quantitative research
generalizability
description and prediction
objective
verifiable
name 4 limitations of quantitative research
context stripping
explanation needed too
value free observation not possible
absolute truth rarely established
what are the key features of quantitative research?
measurable
aims, objectives and hypotheses pre-stated
aim - broad statement
procedures standardised
outcome measures must be reliable and valid
results should be presented statistically
In quantitative research what is an objective?
specific detail (eg. research question)
In quantitative research what is a hypothesis?
predictions of what is expected to happen with a certain intervention
In quantitative research what is a variable?
factors that are being investigated
In quantitative research what is does validity mean?
research must measure what it intends to measure
What should the results in quantitative research aim to be?
falsifiable (hypothesis testing)
establishing causal relationships
establishing association between variables
what is at the top of the hierarchy of evidence in quantitative research?
systems reviews
metanalysis
what is at the bottom of the hierarchy of evidence in quantitative research?
ideas
opinions
editorials
anecdotal
Name the types of studies in order, from lowest to highest, on the hierarchy of evidence
expert opinion case series - case report 1 case series - case report 2 case control studies cohort study 1 cohort study 2 randomised control trial metanylasis / systems review
describe the limitations of an expert opinion
high subjectivity
high bias / errors
little objective research
multiple opinions
describe the features of a case series (case report 1) study
descriptive study
less than 10 participants
performed due to rarity of condition
describe the features of a case series (case report 2) study
more than 10 participants
can’t compare with control group
good for pilot data collection
describe the features of a case control study
observational study conducted in retrospect matched with a control unable to attribute causation control group doesn't have starting condition
describe the features of a cohort study 1
looks at associations between one entity and another - eg. smokers and non-smokers
describe the features of a cohort study 2
data collected prospectively
collect data you want
describe the features of a randomised controlled trial (RCT)
prospective intervention given groups randomly allocated outcome measured and compared very powerful
In a RCT what is an intervention?
something we are doing to the participant
what happens in a RCT when the participants are randomised?
allocated to intervention or control group
In an RCT was is the attention arm?
similar to intervention arm but without active ingredient
what is a null hypothesis?
hypothesises that there will be no difference between the intervention and control arm - study must disprove this assumption
what is a dependent variable?
the outcome of interest (eg. wound healing time) - effect
what is the independent variable?
the intervention factor (eg. wound dressing used) - cause
During RCT recruitment what is the population?
target group we are interested in
What must the target group in RCT be?
clearly defined - gives criteria for inclusion in trial
What is inclusion and exclusion criteria?
included in trial
excluded from trial
what should a sample in a RCT represent?
The population as a whole
why must a sample in a RCT represent the population as a whole?
means data is reliable and generalizable to whole population
what is probability sampling and why is it used?
sampling designed to give unbiased sample
everyone has a chance of selection
name 4 types of probability sampling
simple random
stratified random
cluster
systematic random
what is simple random sampling?
random selection of everyone in population list
what is stratified random sampling?
populations put into groups according to characteristics (gender) and then randomly selected
what is cluster sampling?
random selection of larger units (hospitals) which participants are randomly selected from
what is systematic random sampling?
random selection at predetermined intervals
name 4 factors which affect sample size in RCT
population - similarity of population, expected ‘event ‘ rate, expected attrition rate (drop out rate)
design - higher number variables=larger sample
measure - how sensitive the measure is
practical - cost and convenience
What is a power calculation in RCT?
calculation performed to assess resources needed to power trial
with regard to randomly allocating trial groups in RCT what is simple randomisation?
‘tossing a coin’ usually done with random number tables
with regard to randomly allocating trial groups in RCT why is block randomisation used?
to keep the numbers in each group close
with regard to randomly allocating trial groups in RCT why is stratified randomisation used?
to balance chosen characteristics across the arms of the trial
In RCT what is blinding?
concealing who is in which group (intervention or control)
In RCT what is single blinding?
one person (usually participant) knows which groups they are in, but researcher does not
In RCT what is double blinding?
neither participant or person assessing knows which group they are in
In a RCT what is the primary outcome?
the main outcome of the study
In a RCT what is the secondary outcome?
other outcomes that the study wish to assess, such as cost
What 3 features must outcomes in RCT have to be appropriate?
Meaningful - crediable, meaningful and related to issue of interest
Measurable - amenable to dependable measurement
Responsive - change that occurs is demonstrated in outcome measure
What should measurements used in RCT’s be?
valid and reliable
why is reliability important in RCT’s?
addresses consistency of which an instrument measures attribute it is designed to measure
what does validity in RCT’s do?
proves that the study measurement or assessment does what it intends to do
what is internal validity in a RCT?
degree to which it can be inferred that the independent variable is responsible for the observed outcome
what is external validity in a RCT?
generalizability - shows that results are transferable to other groups
can you achieve external validity without internal validity?
no
name 6 threats to internal validity
history - events happening outside study
maturation - change that happens over time
testing - change due to experience of test
mortality - differences in study drop out
selection bias - participants different to non-participants
name 3 threats to external validity
selection effect - generalizability is affected due to ideal sample population being unavailable
reactive effects - response to just being in study
measurement effects - measurement and testing affects the generalizability
what type of data is primarily collected in quantitative research?
numerical
what are descriptive statistics?
ways of displaying and summarising quantitative data
name 3 ways that numbers are used in data displays
numerical result - eg. BP, age, weight
coded category - eg. 1 =male, 2 =female
ordered categories - eg. 0 =no pain 10 = most pain
name 4 levels of measurement and rank them in order of precision, most to least
ratio
interval
ordinal
norminal
name 4 properties used to define the precision of measurement levels
different categories - all
categories ranked - ordinal, interval, ratio
equal distances between categories - interval, ratio
fixed zero - ratio
what is a strength of displaying data in tables?
data from different variables can be viewed together
what is a strength of displaying data in charts?
immediate visual impact
name the measures of central tendency
mean, median and mode
name the measures of dispersion
range
interquartile range
standard deviation
varience
which levels of measurement be used for measures of dispersion?
interval and ratio
which levels of measurement be used mode?
norminal, ordinal, interval, ratio
which levels of measurement be used median?
ordinal, interval, ratio
which levels of measurement be used mean?
interval, ratio
why do we perform statistical analysis?
to draw inferences from the sample studied about population of interest
name 2 approaches to statistical analysis
hypothesis testing - P values
estimation - using confidence intervals
name the steps involved in hypothesis testing
set null hypothesis set study (alternative) hypothesis carry out significance test obtain test stats compare test stats to hypothesised critical value obtain P value make decision
what is a P value?
probability of obtaining study results if the null hypothesis is TRUE
which numbers does the P value fall within?
0 and 1
which number should the P value be closest to in order to reject the null hypothesis?
0
at what % is statistical significance often set?
5%
what should the P value be to show that there is evidence is present to reject null hypothesis?
less than 0.05
what should the P value be to show that there is insufficient evidence to reject null hypothesis?
more than 0.05
name the 2 types of errors that can occur in hypothesis testing
type 1 - false positive
type 2 - false negative
what is the power of the study?
probability of being able to detect differences in the study groups should one exist
what is the power of the study usually expressed as?
%
what does the % given for the power of the study mean?
if study has 80% power, then there is an 80% chance of detecting a difference between study groups
what are the limitations of using P values?
the only tells us how likely the results are if null hypothesis is true
what are the advantages of using a confidence interval?
give information about effect size and clinical significance of results
what is estimation?
a measure of precision with which the quantity of interest is estimated
what are confidence intervals calculated for?
any estimated quantity from the sample data (eg. mean)
what would a 95% confidence interval indicate?
the range of values within which the true population quantity would fall 95% of the time if study was repeated
in quantitative research which terms refer to internal and external validity?
internal - validity
external - generalisablity
describe a case control study
observational study conducted in retrospect and matched with a control
