2. critical appraisal Flashcards
Research definition
systematic and rigorous process of enquiry which aims to describe phenomena and to develop and test explanatory concepts and theories
- aims to contribute to a scientific body of knowledge
purpose of research
identify or test a theory/ hypothesis
what approach is taken for research
methodological approach
Audit
a count or measurement of current activity/ practice/ performance
- does not address a question or add substantial new knowledge
Evaluation
may involve research methods, or audits
examine either methods or activities
- may lead to new understanding
positivist approach
deductive, - testing through hypothesis development
testing theory - to give explanation, verification, prediction
objective reality - facts
interpretivist approach
inductive - build through empirical examples
developing theory- develop understanding
multiple interpretations of reality - observable symbolic meaning
approach
view of the researcher
overall perspective of the study
methodology
coherent and defined set of methods
methods
practical activity used to achieve the studys aim
scientific method
approach = positivist methodology = quantitative, deductive methods = surveys, experiments, observations
understanding method ??
approach = interpretivist methodology = qualitative, inductive methods = interviews, participants observation, focus groups, document study
healthcare research
focus on treatment
implementation
experience (acceptability, is it a good option)
efficiency - cost and equity
effectiveness - efficacy does it work well
how to find which interventions are effective
experimental design
how to find out what will happen (prognosis) lead to population studies
observational designs
how to find out how things occur or are experienced in a clinical setting
qualitative designs
4 characteristics of answerable questions
PICO
patient or population
intervention or exposure variable
comparison intervention or exposure variable
outcome
to give a testable hypothesis
hypothesis definition
an educated guess or proposition that attempts to explain a set of facts or natural phenomena
T - test
T-test compares the means between two samples of normally distributed data
ANOVA
• ANOVA compares the means between more than two samples of normally distributed data
odds ratio equation
• Odds are calculated by calculating the number of times an event happens by the number of times it
does not happen
• Odds ratios are calculated by dividing the odds of exposure in cases by the odds of exposure in the
control group
Odds of exposure in cases = a/c Odds of exposure in controls = b/d Odds ratio = (a/c) / (b/d)
Odds ratio values
OR of 1 = no difference in risk between the groups
OR > 1 = the rate of the event in experimental group is increased in people who
have been exposed to risk factor
OR < 1 = the rate of the event in experimental group is reduced in people who
have been exposed to risk factor
If confidence interval crosses 1 then the OR is not statistically significant
risk ratios equation
Risk itself is the probability that an event will happen i.e. divide the number of events by the number of
people at risk
• Risk ratio is calculated by dividing the risk in the treated or exposed group by the risk in the control or
unexposed group
risk ratio value
RR of 1 = no difference in risk between the groups
RR > 1 = the rate of the event in experimental group is greater than in control
group
RR < 1 = the rate of the event in experimental group is reduced compared to
control group
If confidence interval crosses 1 then the RR is not statistically significant
absolute risk reduction
Difference between the event rate in the treatment
group to that in the control group.
•ARR allows you to differentiate between
something being statistically significant vs clinically
significant.
number needed to treat
Used to find out how often a treatment works rather than just whether it works
• Number of people who must be treated to result in benefit for one person
Absolute Risk Reduction (undesirable) =
Control Event Rate – Experimental Event Rate
Absolute Risk Reduction (desirable) =
Experimental Event Rate – Control Event Rate
mean
Definition: Sum of all the values, divided by the number of values
when to use the mean
If the spread of data is normally distributed i.e. fairly similar on either side of the
mid-point.
median
Definition: The middle point in the dataset that has half the values above
and half the values below
when to use median
It is used to represent the average when the data are skewed i.e. not symmetrical.
Often given alongside interquartile range (more on this later)
mode
Definition: The most common value within a dataset
when to use the mode
If we need a label for the most frequently occurring event.
Some papers make reference to a ‘bi-modal’ distribution i.e. where there are two
peaks within the dataset
measures of dispersion/ variability - what do they show
❑ Refer to how spread out the data is within a distribution
❑ Can also be called measures of dispersion
❑ Different measures of variability relate to different measures of central tendency
measures of dispersion/ variability - 4 examples
Variance
Standard Deviation
Range
InterquartileRange
Variance
• Definition: The average of the squared differences from the mean
when to use variance
Often as precursor to calculating the standard deviation (more on next slide) to give you an idea of
how spread out your data is from the mean.
❑ Samples with low variance have data that is clustered closely about the mean.
❑ Samples with high variance have data that is clustered far from the mean.
❑ Variance is often used to compare the distribution of two data sets.
Standard Deviation
Definition: The standard deviation measures the spread of the data
about the mean value. It is the square root of the variance.
when to use standard deviation
❑ SD is used for data which are normally distributed
calculating SD and variance
on the slide
Range
Definition: The difference between the maximum and minimum
values in a dataset
Interquartile Range (IQR)
Definition: The difference between the upper and lower quartiles
cohort studies and epidemiology
Cohort studies are of particular value in epidemiology, helping to
build an understanding of what factors increase or decrease the
likelihood of developing disease.
exposure
outcome
confounder
research q = is there a causal relationship btw exposure and outcome
does confounder influence the outcome
is the confounder associated with the exposure
Strengths of cohort studies
•Gather data regarding sequence of events; can assess
causality.
•Examine multiple outcomes for a given exposure.
•Can calculate rates of disease in exposed and unexposed
individuals over time (e.g. incidence, relative risk).
•Observational by nature so participants are not manipulated
in any way.
Weaknesses of cohort studies
You may have to follow large numbers of participants for a
long time.
•They can be very expensive and time consuming.
•They are not good for rare diseases or diseases with a long
latency.
•Differential loss to follow up can introduce selection and
attrition bias
Quantitative Research:
Deductive
Objective
Generalising
Qualitative Research:
Inductive
Subjective
Contextual
• CONVERGENT PARALLEL
- quantitiative and qualitative Data Collection and Analysis
- compare or relate
- interpretation
• EXPLANATORY SEQUENTIAL
- quantitiative data collection and analysis
- data builds to qualitative data collection and analysis
- interpretation
EMBEDDED
- Quant (or Qual) Design
- Quant (or Qual) Data Collection and Analysis
- Qual (or Quant) Data Collection and Analysis
(before, during or after)
= interpretation
MULTIPHASE
- study 1 qualitative
- informs study 2 quantitative
- this informs a third study used mixed methods
4 mixed method study designs
• CONVERGENT PARALLEL - both data used at once
• EXPLANATORY SEQUENTIAL
- quant data informs qual data collection - interpretation
- EMBEDDED
- MULTIPHASE - qual deisgn then quant then mixed methods
