The scientific method:

1
AGRC 3007 Research methodology
The scientific method Definition of the research
problem
Research the process by which new knowledge is
obtained about a natural phenomena
New knowledge how do you know what is new in a
particular area of interest ? If you find one,
how do you know if it is an area worthwhile
researching ?
You are here to learn how to conduct research
effectively, and to become a problem solver.
So in a way, it does not matter what area you do
your research in, what is important is how you
do it, and that you can apply it to solve
problems.
2
It does help if you are interested in a
particular area, as interest is the main factor
that drives your curiosity, or if it is part of
a larger project.
New knowledge how do you know what is new in a
particular area of interest ? If you find one,
how do you know if it is an area worthwhile
researching ?
In this course, to get you going without having
to waste time, you shop around amongst academic
staff and select a project that interest you
starting point.
In environmental science, you have to find your
own projects and supervisors.
In your future job, you will either be given a
broad area to conduct research or you may have
to define your own research areas.
3
That was only the starting point. Only 5 weeks to
go !!!! Whats next ?? How many have made a
decision ?
Familiarise yourself with the general area of the
problem or issue that needs to be resolve, and
prepare yourself to deal with issues around the
problem.
This may appear to be trivial, but it is a very
important step. It will determine how you will
tackle the issues hence the success or failure
of your project.
This step used to be a difficult one when you are
strapped for time, but these days it should not
be a limiting step at all. Literature search is
now very easy. The session from the library on
databases and endnote was intended to assist you
in this aspect. If you missed this catch up !!
4

• Search the literature through the library
• Issues directly associated with the problem
• Issues related to the problem
• Methodologies used in this area of research
• Different opinions on the subject
• Most importantly summarised it in your mind and
• synthesize the ideas into your own.

2. Talk to your supervisors or to their post
graduate students. They are generally ahead
of you and you should pick their brains, and
move further from there. You dont have to make
the same mistakes they made, you should add to
their collective knowledge.
3. Talk to other scientists/consultants/farmers
that are having to deal with the issues of
interest.
5
You have a problem or issue to do your research
project
You also have done your critical literature
review, you have tap the supervisors, the post
graduate students, other people.
You need to match these. You need to use the
literature to justify your research proposal.
The What, Where, When, Why and How.
If it does not match, you will need to revisit
the issue you selected or that was presented to
you (the hypothesis) and perhaps rethink a new
hypothesis. This should be done in consultation
with your supervisors.
An essential part of this work should be the
decision on the fundamental approach of how to
solve the problem
6
The scientific method Application of the
scientific method operationalising the approach
adopted.
Develop your Materials and Methods
proposal This should be backed by your literature
review. A good knowledge of the various methods
that can be adopted will be very useful if the
method adopted does not work for you during the
project execution. You will need to be able to
change your experimental details and such
knowledge will assist you in making decisions,
particularly if you are 500 km away in the
field. KNOWLEDGE INDEPENDENCE
7
METHODOLOGY DATA COLLECTION
Each area of research has its own particular
pitfalls to be avoided, but it is extremely
important to exercise maximum care and skills at
this stage of a project Reliability of your
conclusion can never be better than the
reliability of the data on which you base your
conclusion.
Example of common pitfalls Chemistry some
reagents can loose their effectiveness with age
results will be dependent on the reagent
age. How would you overcome this ?
Use fresh reagents every time expensive !!!
Use a standard or reference sample (soil or other)
8
Another example Working with pastures factors
sffecting botanical comp. All, samples came into
a lab where several lab assistants sorts them
into various species and weigh them on a
platform balance. Large project, collected a
large amount of data routinely, hence no
supervision required anymore. One day, one of the
research scientist went through the data and
noticed that over a period several samples weigh
exactly 2.1 kg and none were above this. He
investigated.
Found a coffee cup stuck under the balance and
no-one noticed This cup held up the platform at
2.1 kg. Data was useless.
9
My own experience On a 6 months sabbatical,
working with salt accumulation in the free space
of leaves. Had access to a new modern
Perkin-Elmer carbon-arc AAS I volunteered to set
it up and had it going well and calibrated It.
After a while I noticed that the calibration
changed from a convex to a concave shape. Checked
and rechecked, and decided to proceed because my
samples were well within the range of
calibration. Conducted 8 - 10 experiments.
10
Interpretation of data collected
First step is to tabulate the data, finding out
which are statistically different and which are
not.
Next step is to experiment different ways of
plotting/presenting the data and get an idea of
the trends of treatment effects. Different ways
of graphing may show different effects scale !!
Next important step is to put your head on the
chopping block, saying what you think the
results mean, and what action should be taken as
a result of them
It is not acceptable to simply say here is the
data I collected.. .. I am not saying what it
means Its up to you to work it out
11
Remember that you decide which relationships and
treatment effects are worth testing. It is not
the biometricians or the computers job to find
out what your data means, they help you to test
the validity of what your conclusions should be.
Statistically significant differences may not
mean much, its the magnitude of these changes
that are important.
Do not forget the possibility of interactions
between treatments.
12
REPORTING OF RESULTS
This is an extremely important part of any
research. No matter how well the investigation
was conducted, it is useless if the findings is
not made known to people who can make use of the
information.
Many reasons for not reporting, eg research is
continuing, not confident with the results,
writing is hard-research is more enjoyable.
These days money is allocated more for applied
research. hence reporting results is very
important for continuity.
Because the researcher and the end-user do not
know each other, a fundamental requirement in
reporting is HONESTY. It is a fundamental basis
of modern science.
13
Standards of scholarly integrity.
Total honesty moral obligation
Cheating is regarded very seriously by University
and the scientific community, which may take the
form of
Plagiarism presenting ideas or material from
someone elses work as your own.
Selective reporting including observations that
fits a particular hypothesis, but excluding
those which did not.
Falsification altering experimental results or
making up false exptl results that fits into a
particular hypothesis.