1)
Why experiments need to be designed
-
Very simple, need common sense,
biological insight and careful planning
-
Myth 1: It does not matter how you
collect your data, there will always be a statistical way that will allow you
to analyze them (Independent data points and other assumptions)
-
Myth 2: If you collect lots of data
something interesting will come out, and you will be able to detect even very
subtle effects (quantity of data is no substitute for quality)
2)
The cost of poor design
-
Time and money (limited return if any
on the efforts and resources invested)
-
Ethical issues (I can not emphasize
too strongly that while wasting time and energy on badly designed experiments
is foolish, causing more human or animal suffering or more disturbance to an
ecosystem than is absolutely necessary is inexcusable. They can be avoided by
careful consideration of how your experiment is designed)
3)
Relationship between experimental
design and statistics
-
It is essential that you think about
statistics that you will use to analyze your data before collecting them
-
Every statistical test will have
slightly different assumptions about the sort of data they require or the sort
of hypothesis that they can test.
-
Designing experiments is as much
about learning to think scientifically as it is about the mechanics of the
statistics that are used to analyze the data. It is about having confidence in
your data
-
Experimental design is about the
biology of the system, and that is why the best people to devise biological
experiments are biologists themselves
4)
Why correct experimental design is
important to biologists
-
The two major goals of designing
experiments are to minimize random variation and account for confounding
(extrinsic) factors.
-
Random variation (Between-individual
variation, within-treatment variation or noise) quantifies the extent to which
individuals in our sample differ from each other.
-
Random variation is everywhere in
biology but things are different in other branches of science (physics and
weight of electrons).
-
Confounding (extrinsic) factors: If
we want to understand the effect of one factor (A) on another factor (B),
however, our ability to do this can be undermined if B is also influenced by
another factor (C) which is the confounding factor
-
Example: Effect of sunlight on the
weight of a certain fish when temperature is closely linked
to sunlight and also affect the weight of the fish.
Summary
1)
We cannot be a good biologists
(biotechnologists) without understanding the basics of experimental design
2)
The basics of experimental design
amount to a small number of simple rules, without the need for complex
mathematics
3)
If you design poor experiment, then
you will pay in time and resources wasted
4)
More important than time and
resources is the need to reduce suffering to animals or humans or disturbance
to an ecosystem