Potential Sources of Variability
|
Questions to Ask
|
Points to Consider
|
Viability (health) of the
seeds
|
What percentage of the seeds will sprout under ideal conditions?
|
If fewer than 80% sprout in your control, you may have a
problem with your seeds or growing conditions.
|
Definition of "germination"
|
Does everyone agree on how germination is defined? If a
seed sprouts but has no distinct radicle that can be measured, do you
count it as having germinated or not?
|
This is the kind of decision that each experimenter can
make, but in order to compare the results of your experiments with those
of other scientists, you want to clearly state what decisions you have
made.
|
Precision of measurement
|
If several people measure the same radicle, do they come
up with exactly the same measurement? Is there greater variability in
the data if several people take the measurements, compared with having
them all done by one person?
|
Judgement is often important in scientific measurements.
At what exact point do you start measuring the radicle, and how tightly
do you stretch it while measuring its length? Would everyone make the
same decisions? Would you do it the same way every time?
|
Bias
|
Are you tempted to choose the best looking seeds for your
control, and smaller seeds for treatments that you don't expect to grow
much anyway? If you knew you would get a higher grade if your data indicated
clear differences between the control and the treatments, would this
affect how you selected which seeds to put in each dish?
|
Scientists continually have to be on the lookout for sources
of bias in their experiments. They are much more likely to get their
results published and to continue receiving funding if they get good
results from their experiments. Why do you think scientists might not
want to be biased, in spite of these pressures?
|