School

University of New South Wales **We aren't endorsed by this school

Course

VISN 3211

Subject

Statistics

Date

Aug 7, 2023

Pages

17

Uploaded by mrstwitch on coursehero.com

EBP AND REVIEW OF
STATISTICAL MODELS
Evidence based practice
- combination of the best available research, the patient's
preferences/circumstances, clinical environment and practitioner's expertise
Five steps in EBP:
1.
Asking clinical questions - translation of uncertainty to an answerable question
2.
Acquiring information - systematic retrieval of best evidence available
3.
Appraising information - critical appraisal of evidence for validity, clinical relevance
and applicability
4.
Applying information - application of results in practice
5.
Auditing practice - evaluation of performance
Appraisal
- evaluating the relevant research evidence to find the highest quality evidence
available relevant to your question
Critical appraisal
- process of assessing and interpreting evidence by systematically
considering its validity and its relevance to the question
Internal validity
- extent to which the research is reliable
External validity
- indication of the generalizability of the findings
Considerations and codes for research ethics
1.
Human subject protection - minimizing harm, respect human dignity and privacy
2.
Honesty - reporting findings, interpretation and publication status, deception and
misrepresenting data, fabrication, falsification or plagiarism
3.
Objectivity - avoiding bias in experimental design, commercial interests, data
interpretation
4.
Openness - share data and results, open to criticism, resources
5.
Confidentiality - protect identification of subjects and records
How to ensure standards
University level
Organisations/committees
Publishing bodies
Funding bodies - Australian Research Council
Common law
Correlation
- way of measuring the extent to which two variables are related
Measures the pattern of responses across variables
Observing what naturally goes on in the world without directly interfering with it
Varies between -1 and +1, where 0 = no relationship
Effect size:
0.1 = small effect,
0.3 = medium effect,
0.5 = large effect
Page 1 of 17

Coefficient of determination, r
2
by squaring the value of r you get the proportion of
variance in one variable shared by the other
What to consider
The significance of r
The magnitude of r
The +/- sign of r
For example
There was no significant relationship between
the number of adverts watched and the number
of packets of toffee purchased, r = 0.87, p
0.54
r = 0.87 is a large effect
the sign of r is positive - as one variable
increases so does too the other but this does not
imply causation
Describing a straight line
Y
i
=
b
0
+
b
1
X
i
+
ε
i
b
1
- regression coefficient for the predictor, gradient of the regression line,
direction/strength of relationship
b
o
- intercept, point at which the regression line crosses the Y-axis
Summary of linear regression
Way of predicting one variable from another by fitting a statistical model to the data
in the form of a straight line which best summarises the pattern of data
We have to assess how well the line fits the data using:
o
R squares - tells us how much variance is explained by the model compared
to how much variance there is to explain in the first place, proportion of
variance in the outcome variable that is shared by the predictor variable
Page 2 of 17

o
F which tells us how much variability the model can explain relative to how
much it can't explain i.e. ratio of hwo good the model is compared to how
bad the model is
o
B-value - tells us the gradient of the regression line and the strength of the
relationship between a predictor and the outcome variable, if its significant
(sig. < 0.05) then the predictor variable significantly predicts the outcome
variable
t-test
Independent t-test
Compares two means based on independent data e.g. data from different groups of
people
Dependent t-test
Compares two means based on related data
E.g. data from the same people measured at different times
Data from matched samples
observed difference betweensample means
−
expected difference
between populationmeansif nullhypothesis istrue
¿
estimateof the standard error of thedifference betweentwo samplemeans
t
=
¿
i.e. we compare the difference between the sample means that we collected to the
difference between the sample means that we would expect to obtain if there was no effect
Use the standard error as a gauge of the variability between sample means
If the difference between the samples we have collected is larger than what we
would expect based on the standard error then we can assume one of the two:
o
There is no effect and sample means in our population fluctuate a lot and we
have, by chance, collected two samples that are atypical of the population
from which they came
o
The two samples come from different populations but are typical of their
respective parent population. In this scenario, the difference between
samples represents a genuine difference between the samples (so the null
hypothesis is incorrect)
Type I error
- occurs when we believe that there is a
genuine effect in our population when in fact there isn't,
probability is at
-level (usually 0.05)
Type II error
- occurs when we believe that there is no
effect in the population when, in reality, there is, probability
is the
-level (often 0.2)
Non-Parametric tests Lecture Outline
Page 3 of 17

Page1of 17