Your assignment was to design an experiment to
determine which of two lakes would be better suited for the largemouth
bass fishing contest desired by your client. Lakes Acorn and Bacon are
similar in size (about 300 ha), age and physical structure. The decision,
then, will be based upon the bass populations in each lake and their accessibility
to anglers. Since the lakes are publicly owned, it will not be possible
to drain the lakes, or to otherwise kill the fish (which, by the way, would
make for a lousy fishing contest). Electronic counting cannot be counted
on to provide sufficiently accurate measures (nor would it necessarily
reflect which lake's fish are easier to catch). As a result, your best
bet is to simulate the contest in both lakes and recognize that the bass
caught by your fishermen are samples. These two samples can be used to
compare the population size and accessibility for each lake.
There are a number of decisions you must make
about how to simulate the fishing contest. These operational definitions
include WHO fishes (since humans differ in skill level, knowledge, and
motivation with respect to fishing), WHEN they fish (fish accessibility
may vary seasonally and with time of day), HOW they fish (not all fishermen
use the same techniques) and even WHAT constitutes fishing (hook and line,
fish stunners, dynamite...).
Some of these decisions ought to be easy to make.
Since the fishing contest will be part of a family reunion in early June,
sampling should take place as close as possible to the time of the reunion.
Your fishermen should fish during the same hours as the contest (6am -
8pm) and follow the "hook and line only" rule that will be followed by
the contestants. As far as the "WHO", it makes no sense to use people who
either don't know how to fish or do not want to fish. Use experienced bass
fishermen and the more the better (unless you have so many fishermen that
they deplete the bass populations). A larger sample of fishermen will increase
the likelihood that your sample will reflect fishing methods/techniques
of the contestants. These fishermen should not be familiar with either
of the two lakes. They should be told to inspect the lakes and use their
skill and knowledge to catch as many fish and the biggest fish possible.
Even though the winner of the contest will be the one with the biggest
fish, contestants will have more fun if they catch some fish. Once these
decisions are made, designing the experiment is straightforward. For instance:
Ten experienced bass fishermen will be chosen
to sample bass populations on each lake. Each fisherman will fish a ten
hour day on each lake during the last week in May. Fishermen will use their
own equipment and techniques, but will be instructed to fish the same way
on both lakes. None of the fishermen used will have fished either lake
before the experiment begins. All bass caught will be counted and the weights
recorded for fish over one pound.
(Notice also that the experimental design outlines
what data will be gathered and implicitly how these data will be analyzed.)