Project: The project is your independent research and data analysis based on either primary (you collect your own) data or use some existing data. You are to design the study, collect the data (or use some existing data), analyze the data, draw appropriate conclusions, and document your study in a carefully written report.
?It is important that you decide on an objective (ie. some question of interest that can be studied with a statistical investigation) and then?find or generate the appropriate data to answer the question. Finding a nice clean data set and then making up a question is not recommended and will seriously impact your grade. ?Use of data sets that accompany the current textbook or other books should not be used. Use of “canned” data ?sets from the internet such as KAGGLE, ?UCI Machine Learning , etc are not recommended. Your grade will consider the originality of the data.?You may lose up to 3 points (out of 25) for using one of those data sets.
?Reports or papers are preferred over power-point presentations, posters or other forms of presentation. ?Computer files are not acceptable. See below for some guidelines on the organization of the final report.
One possible plan (but, not necessarily the only one) Start with some hypothesis, conjecture or claim. The claim might be yours or someone else's (for example: a manufacturer that claims their product lasts longer than that of a competitor). Figure out the best way to test this claim. The first step might be to figure out the correct response(s) to measure (ordinal, nominal, interval, ratio, discrete, continuous). Decide on a plan of attack, which might include doing an observational study or a designed experiment. The data to put your claim to a test might be available somewhere (on the Internet?) or you may have to collect it yourself. If you collect the data yourself, decide on an appropriate method of collection (random sample, cluster sample, etc.) and an appropriate sample size. Justify the method you used and the sample size selected. Determine the appropriate analyses to test the claim. Use of descriptive statistics is a good start but use of statistical inference (confidence intervals, hypothesis testing, etc.) should be considered a requirement. Use of more advanced analyses and specifically the techniques covered in this class should be considered. Use of elementary statistics (like t-tests, chi-square, simple regression, etc) might be an indication that your project idea lacks sufficient complexity to be considered for a high grade. Carefully complete the analysis. Draw an appropriate conclusion and carefully document what you did and your conclusions. It may also be appropriate to comment on lessons learned and potential improvements if the study was to be repeated (what you might do differently) Projects will be ranked using the criteria shown below. The instructor reserves the right to grant extra credit for a truly exceptional project (rarely done) and to give a failing grade to extremely poor projects.
Projects should be done independently. Projects involving collaboration between two or more people will NOT be accepted this semester.
Criteria for a "good" Project