The goal of this project is to collect and analyze your own data using
the methods you have been learning this quarter. In
particular, you need at least 3 variables, at least 2 quantitative and
at least one categorical. You are advised to collect multiple variables so you have some flexibility in which to
analyze. You should specify a different, fairly
specific research questions involving 2 or more explanatory variables
(quantitative and/or categorical). You may use either a quantitative or
categorical response variable (though we won't talk
about categorical response variables until the end of the quarter). You may use
different response variables for different questions but don't
have to. The project will be 15% of your course grade.
It is up to you whether you conduct an observational study
(with random sampling) or a randomized experiment (which are often easier to
implement). You should select a research question that is either related to
your thesis work or is interesting to you. You should collect your own data in
a manner appropriate for the analysis. If it is impossible to collect the data
appropriately, collect it in a convenient way, making certain that you (1)
identify the appropriate method for collection, (2) explain why you deviated
from the ideal method, and (3) what the possible consequences are using your
methods versus the ideal. You may also analyze data collected by a research
partner or faculty advisor. In your final report, be sure to explain how,
where, and when you collected the data, and any potential measurement problems
with the data.
This proposal should outline your preliminary research questions (e.g.,
observational/experimental units, sample(s), population of interest, variables,
classify each as quantitative or categorical) and your plan for how you have
collected (or will collect) the data and how you included randomness in the
study design. These are tentative proposals but will make sure you have begun
thinking about the project. You will receive peer feedback on your
proposal. Be prepared to make modifications to your topic after this feedback.
(This is also to allow me to make sure your analysis is within the scope of the
methods we will learn in this course.)
Final report
You must submit a final report by the final exam. The report
will consist of two parts: (1) a 1-3 page executive
summary and (2) a technical write-up. Reports must be typed, with relevant
computer output incorporated into the body of the paper. Also submit a copy of
your raw data file.
The executive
summary should be professionally formatted and written for
a non-statistician. It should be similar to what you
might read in journal from your field.
It should...
·
describe the
research question
·
summarize how
the data was collected
·
summarize the
data in tabular, graphical, or other method appropriate for effective communication
·
clearly
identify the primary analysis methods and variables used
·
present
conclusions and implications (i.e., answer the research question). Cite
test-statistic values, df, and p-values used.
The technical write-up should
be typed and well organized. The technical write up should document your
analysis methods. It should include snippets of JMP output (including
diagnostic plots) used to make important analysis decisions and a general
outline of how you went about analyzing your data. Someone who has taken our
class should be able to look at the technical write-up and decide whether your
procedures were appropriate and thorough for the claims being made in the
executive summary. Also, any additional technical discussions of the data
should appear in this section. There is no limit to the amount of output or
graphics in this section, but your commentary should not exceed 2-3 pages. Note
that you do not need to include every piece of output you used, but you should
include and reference important items that helped you make decisions along the
way.
Project Presentations
You will make a presentation of your project for the rest of the class. You will use PowerPoint or some other presentation software. The presentation will be at most five minutes (this is harder than it sounds – you should practice the timing before hand!). Feel free to be creative. We will either have in class presentations or recorded presentations posted online. More details to follow.
Your grade
will be determined in part by:
·
the quality of
your proposal
·
how well you
described your research question
·
the
appropriateness of your analysis
·
the
completeness/thoroughness of your analysis
·
your
communication of the results, both in the executive and technical portions.
Remember to include the executive summary, technical
support, and data file as separate files.
Your report must be word processed with all relevant computer output incorporated into the body of the report. (Graphs and output should be integrated into the report, not just as an appendix. You can include additional details as an appendix but still need to select the relevant information as part of the discussion.) You should again include the original data worksheet when you upload your report.