Transcript How do I get involved?

An Introduction to
Quantitative Research
Peter Harper
Quantitative Research is dominated by two approaches
The Experiment and The Survey
The Experimental Method
William Harvey (1578 - 1657)
Harvey observed that blood
moved around the body in a
circle, propelled by the heart
out through the arteries and
back through the veins.
Viewed in the broadest terms,
the value of the experimental
method is demonstrated.
Francesco Redi (1626 -1697)
In the 1600s, it was widely believed that
living things could spontaneously generate
from non-living, dead, or waste materials
because people saw such materials
"generate" living things such as mould or
maggots.
In 1688 the Italian naturalist Francesco Redi
set out to test the idea that flies were
responsible for the maggots by placing
decaying meat in two containers: one open
to the air, the other sealed. Meat in the
open container eventually became infested
with maggots.
Francesco Redi (1626 -1697)
In another experiment Redi drank
Viper venom to show that it was
only dangerous if injected directly
into the bloodstream
Isaac Newton (1636 -1727)
People were using prisms to
experiment with colour, and thought
that somehow the prism coloured the
light. Newton obtained a prism, and set
up his so that a spot of sunlight fell
onto it.
In their experiments, Descartes, Robert
Hooke and Edward Boyle had put a
screen close to the other side of the
prism and seen that the spot of light
came out as a mixture of colour.
Newton realised that to get a proper
spectrum you needed to move the
screen a lot further away.
Isaac Newton (1636 -1727)
In the study upstairs at Woolsthorpe,
he used the 22 feet from the window to
the far wall to project a beautiful
spectrum.
The white light split into different
colours and each colour had been bent
a different amount by the prism.
But to prove that the prism was not
colouring the light, Newton did an
Experimentum Crucis - his crucial
experiment. He put a screen in the way
of his spectrum, and this screen had a
slit cut in it, and only let the green light
go through.
Isaac Newton (1636 -1727)
Then he put a second prism in the
green light. If it was the prism that
was colouring the light, the green
should come out a different colour.
The pure green light remained green,
unaffected by the prism. Newton had
proved that white light was made up
of colours mixed together, and the
prism merely separated them - he
was the first person to understand
the rainbow.
Are there methodological differences between the
“experiments” undertaken by Harvey and Redi?
Inductive Reasoning
Harvey’s experiment illustrates an inductive
approach to scientific method.
Inductive reasoning is based on meticulous
observation, e.g., we can observe that metal
expands when it is heated.
Deductive Reasoning
Redi’s experiment illustrates a deductive approach to
scientific method.
Deductive reasoning is based on prediction, e.g., we
can predict that if we don’t leave periodic gaps in
railway tracks they will buckle.
Hypotheses
Central to deductive reasoning is the hypothesis.
HYPOTHESIS: a suggested explanation for a groups of facts
or phenomena either accepted as a basis of further
verification or accepted as likely to be true.
(Collins English Dictionary 2000)
Using this definition the hypothesis can be seen as either:
•
the purpose of an experiment (testing the hypothesis),
e.g., Francesco Redi set out to test the hypothesis that
life could spontaneously occur; or,
•
the product of an experiment (generating the hypothesis),
e.g., William Harvey
Experimental Method
hypothesise
do something
observe
hypothesise
do something
(Intervene)
observe
(measure)
measure
intervene
measure
hypothesise
Experimental Method
experimental group
hypothesise measure
intervene
measure
control group
hypothesise measure
don’t intervene measure
experimental group 1
hypothesise
measure intervene
measure
experimental group 2
hypothesise
measure intervene
measure
control group
hypothesise
measure don’t intervene measure
experimental group 1
hypothesise
measure intervene
measure
experimental group 2
hypothesise
measure intervene
measure
experimental group 3
hypothesise
measure intervene
measure
control group
hypothesise
measure don’t intervene measure
Variables
Intervention
The independent
variable
may or may
not cause a
change in
The independent variable and any
associated measurements must be
reliable and valid.
In order to be sure that any change in
the dependant variable is the direct
result of the independent variable the
extraneous variables have to be
identified and adequately controlled.
Outcome
The dependent
variable.
The dependant
variable may be
influenced by one
or more
extraneous
(confounding)
variables.
Types of Quantitative Design
A Randomised Controlled Trial (RCT)
A randomised controlled study is one in which:
There are two groups, one intervention group and one control group.
The intervention group receives the treatment under investigation, and the control
group receives either no treatment or some standard default treatment.
Patients are randomly assigned to all groups.
Types of Quantitative Design
A Quasi Experiment
“A quasi-experimental design is one that looks a bit like an
experimental design but lacks the key ingredient -- random assignment.
My mentor, Don Campbell, often referred to them as "queasy"
experiments because they give the experimental purists a queasy
feeling. With respect to internal validity, they often appear to be inferior
to randomized experiments. But there is something compelling about
these designs; taken as a group, they are easily more frequently
implemented than their randomized cousins.”
Trochim WM. The Research Methods Knowledge Base, 2nd Edition.
Available at: http://trochim.human.cornell.edu/kb/index.htm
(version current as of August 16, 2004).
Types of Quantitative Design
A Case Control Study
Case control studies are studies in which people who are already
in a certain condition are compared with people who are not.
Types of Quantitative Design
A Cohort Study
(Longitudinal)
A Cohort Study is a study in which people who presently are in a certain condition
and/or who receive a particular intervention are followed over time and compared
with another group who are not affected by the condition under investigation.
Types of Quantitative Design
A Typical Pyramid of Evidence
Source = http://servers.medlib.hscbklyn.edu/ebm/2toc.htm
Systematic Review
A systematic review is a comprehensive review of a topic in
which all of the primary studies of the highest level of
evidence have been systematically identified, appraised and
then summarised according to an explicit and reproducible
methodology.
A systematic review may or may not include a meta-analysis.
A Meta-Analysis
A meta-analysis is a review in which the results of all of the included studies are
similar enough statistically that the results are combined and analyzed as if it was
one study. In general a good systematic review or meta-analysis will be a better
guide to practice than an individual article.
Survey Method
The most common way of collecting survey
data is via a questionnaire.
Questionnaires can include both open and
closed questions.
Survey Method
Questionnaires can be completed on paper
(eg. post), electronically (eg. internet) or
verbally (eg. by telephone)
Questionnaires are usually analysed
statistically. Answers to open questions are
quantified.
Sampling for Experiments
People
Groups of what?
Things
Population = the total number of
people or things available
experimental group
hypothesise measure
intervene
control group
hypothesise measure
don’t intervene measure
Where do we get
these groups from?
The whole population
A sample taken from the population
measure
Sampling for Surveys
Group of what?
survey group
Where do we get
these groups from?
People
Population = the total number of
people available
Complete survey questionnaire
The whole population
A sample taken from the population
Population and Sampling
Total or Theoretical Population
Target Population
Sample
Probability Sampling
(Random)
Non-Probability Sampling
(Non- Random)
Probability Sampling
Simple Random
Suitable for homogenous populations each
unit should have an equal chance of being
chosen.
Stratified Random
Suitable for populations which vary greatly
in terms of age, gender, education,
experience, illness condition.
Systematic Random
Units are chosen from the sample frame list
at predetermined chosen intervals starting
from a randomly chosen point.
Cluster Random
Suitable for populations which are arranged
in clusters e.g.. Hospitals. Clusters are
sampled first followed by units within
clusters e.g.. Wards.
Population and Sampling
What are the characteristics of an ideal
sample for quantitative research?
The sample is
randomly
selected and
if grouped,
randomly
allocated
The size of the
sample reflects the
population diversity
and the study
variables
The sample is
representative of
the population
Statistical Analysis
There are two main types of statistics
DESCRIPTIVE and INFERENTIAL
Experiments are usually analysed using inferential statistics
in which statistical significance is inferred. i.e., that the
difference observed between two or more variables did not
happen by chance alone
Surveys are often analysed using both descriptive and
inferential statistics
Generalisation
The ultimate purpose of quantitative research is to be able to generalise the
outcome/s to the population from which the sample was drawn.
In order to be able to generalise the outcomes the research must have:
Survey & Experiment
a randomised,
representative
sample of an
adequate size
Experiment
Experiment
reliable and valid
interventions and
associated
measurements
adequate
identification and
control of
extraneous
variables
Thank you
Any Questions