Biology Reference
In-Depth Information
factor(s) behind secular change affecting the timing of epiphyseal fusion in humans?
This question is much more complex and would require a number of tests to answer it
properly.
At this stage, you want to state your question in specific terms and you will further refine
this into your hypothesis later. For example, don't ask, “What can lumbar vertebrae reveal
about an individual?” Rather, ask, “Can the fifth lumbar vertebra (L5) be used to estimate
stature?” You will base your question on the observation(s) you have already made and
ask if it is testable given current knowledge and technology. If your question is whether or
not the fifth lumbar vertebra can be used to estimate stature, then yes, that is a testable ques-
tion. You could measure the height of several L5 bodies from a sample of individuals for
whom you have living stature information. You could then perform a regression analysis
to determine if the height of the vertebra correlates with living stature (refer to Moore and
Ross [Chapter 6], this volume for more information on stature research).
If your question is technically testable but the test would require logistic planning and
resources that would be prohibitive, then you will want to pose a different question. For
example, if you wish to develop a stature equation based on the femur for a sample of
6000 skeletons but you will be working by yourself and you only have one week to do it,
you will have to accept a smaller sample size or come up with a different project. Once you
have determined whether or not your question is testable, begin thinking about how you will
go about designing the test.
At this early stage, most likely all you have actually done is think about your observation
and problem. What you must also do is background research on your problem and question
(refer to section on library research in DiGangi [Chapter 17], this volume; and the upcoming
section on developing ideas, this chapter). There is no point in reinventing the wheel if
research has already been done on the exact same question. However, reasons to repeat
a study might include testing different statistical techniques, to verify the results, to improve
upon a method, or perhaps to validate a particular method for another population. Even if
the same exact question has not been studied before, it is likely that similar questions have
been, and you can use these studies to help you formulate your own research design.
Science builds on what has been already done
d
there is no reason to come up with an
entirely novel research design unless it is your goal to do so. Do not skip over this back-
ground research step. We attended a presentation once in which the researcher presented
their hypothesis for unusual features found on some skulls. Unfortunately, it turned out
that these unusual features were simply a normal anatomical variant, a fact that was pointed
out to the person at the conclusion of their presentation. Doing library research initially after
having first made the observation of the features would have saved them a lot of time and
embarrassment.
Step 3: Hypothesis Testing
As outlined earlier, a hypothesis is a scientific statement based on an observation(s) that is
both testable and falsifiable. Hypotheses that are valid meet these two criteria. This is impor-
tant because as mentioned previously, science works by testing and disproving questions.
For example, a statement such as, “osteophyte formation on the lumbar vertebrae is age-
related” is both testable and falsifiable. You could use a sample of skeletons that have real-age
