You will pick a topic of interest to you in which there is a hypothesis you wish to test, design an experiment to test this hypothesis, and implement a pretest of your experiment. Our Hanover tap water experiment could serve as a template (though we did it in the wrong order! ). In the process of collecting data and formulating the project you may work in groups; however, everything you turn in must be entirely in your own words, and you must carefully reference those who helped you design and implement your ideas and describe the nature of their contributions. There are two parts to be turned in:
(a). (PROPOSAL) Your will clearly explain your your proposed implementation and as well as any other pertinent facts about your experiment in a 2-4 page paper (double spaced) prior to the experiment. In this paper you should discuss your goal, protocol, choice of test statistic, and choice of significance level as precisely as possible.
Due Date: Wednesday May 19. (Here is a sample
proposal
and follow
up).
Comment: Please keep your test as "innocuous" in nature as
possible. The official word on hypothesis tests involving Human Subjects
at Dartmouth is as follows: Dartmouth College has an office on campus
referred to as the office of the Committee for the Protection of Human
Subjects (CPHS). This office is federally mandated to review all research
involving human participants at the College (and associated institutions).
A research study will receive one of three levels of review: Exempt,
Expedited or Full Committee review. The general scope of the project for
this course has been approved with a designation of Exempt by the CPHS
office because the information you plan to obtain is considered
"innocuous" in nature. Your professor will be reviewing your particular
project. It is possible your project may not fall into the exempt
category, in which case your professor will contact the CPHS office. This
is particularly true if your research involves minors or if it involves
information that could place a participant at risk of criminal or civil
liability or be damaging to the subject's financial standing,
employability, or reputation. For our project, I highly recommend that
you do not involve minors.
Also, when designing your experiment please keep in mind the following
guidelines: Prior to asking a participant to become involved in a research
study s/he should be aware of: your name and affiliation with Dartmouth,
the reason for the project, the level of confidentiality of responses, and
the voluntary nature of the person's participation. This may be
accomplished verbally or through a few sentences at the beginning of a
survey instrument (see information sheet template in the "Forms" section
of our web-site: www.dartmouth.edu/~cphs)
(b) (CONCLUSIONS) After running the experiment an analysis of
your results
will be performed and turned in as a second 2+ page (double spaced) paper.
This paper will include a presentation of your results, an analysis of
your results, and a conclusion concerning your goals. Also you should
include a description of any problems that arose while implementing your
experiment and a discussion of any improvements you would implement if
you were to run a second more comprehensive pretest.
Due Date: Friday May 28.
(Here is a sample
analysis, though it does NOT include the graph files.)
Some questions that I will keep in mind when grading your Proposal and
Conclusion are:
1. Are the notions of double blind, controlled and randomized appropriate
to your experiment and implemented sensible way?
2. What was your method of choosing a sample population, and how did you
test to see whether your sample population was an appropriate sample of
the population that you made inferences about?
3. Does your project use the correct language and calculations for
conducting a hypothesis testing?
4. Was your choice of test statistic relevant to your goal, well thought
out, and analyzed correctly?