Biology Reference
In-Depth Information
FIGURE 2.1 Hypothesis testing flowchart. Note that no matter the outcome of a test, new hypotheses are
generated based on the critical analysis of the results.
identified the skeletal collection you wish to use for your project. Logistics with regards to
access, permissions, travel, etc., all become important here. (Refer to the upcoming section
on developing a research idea for more information.) Assuming that you have already
worked all these issues out, you will need to sit down and figure out several things: (1)
data collection method; (2) appropriate sample size; (3) possible sources of bias inherent to
the sample; (4) how to contend with intra- or interobserver error; and (5) how to analyze
the data. Each of these deserves further mention.
A good discussion of sample size and intra- and interobserver error as it relates to skeletal
biology can be found in Buikstra and Ubelaker (1994) . Consult this reference for a complete
discussion; however, to summarize you need to conduct a statistical analysis of intraobserver
error (if doing all the scoring/measuring by yourself) and/or a statistical analysis of interob-
server error (if two or more individuals are scoring/measuring). This is to demonstrate the
replicability of your study. If either you or between two or more observers are not measuring
or scoring consistently, then this will negatively affect the results and introduce a lot of error.
Ideally, your study will have a very low level of this type of error. See Smith (Chapter 7), this
volume for more information on intra- and interobserver error.
Further, you need to consider the introduction of unintentional bias on the part of each
observer. For example, if you are conducting an aging study but the real age of the individual
Search WWH ::




Custom Search