Professor Lesser’s Stat 1380 Data Analysis Project guidelines
1.) In order for effective cooperative learning and oral presentation scheduling, each project must be done in a team of 2-4 people (unless special permission is granted by the instructor), so start thinking early in the semester about choosing teammates with compatible schedules, interests, working styles, etc., because learning to work well in teams is required for virtually every profession. Be sure you know how to reach each other and maintain backup copies of your work (whether GoogleDocs, Dropbox, GoogleDrive, etc.).
2.) Each team will present its project (and turn in a single written write-up for the team as described below) on the meeting date specified in the syllabus and confirmed by the instructor. If you have a pre-approved reason not to be able to present in person during this time, you need to arrange with me as soon as possible to present (or, if necessary, upload as a YouTube video) your project earlier. It will take more time than you think to do a thorough, high-quality job, so do not try doing most of the work in the last 10 days.
3.) An ORAL PRESENTATION is a required part of the project, but is assessed in such a way so that it can only help your grade if your final average falls a point short of a letter grade cutoff. You may find it helpful to check out the “tips for oral presentations” webpage. I’ll confirm later the exact amount of time each team will have (e.g., it depends on how many teams there are), but you’ll probably have only about 5 minutes for your presentation followed by 1 minute for discussion/questions as the next team starts getting set up. Obviously, 5 minutes is enough only for highlights, not every detail in the written writeup you’re turning in to me. For the presentation part, you’ll have access to the technology in the room (but make sure things are working and “ready to go” so that you don’t use up part of your time just “setting up” or materials or booting up software/websites) and any additional manipulatives or materials you bring. Each team member must have a major non-silent role in the presentation. Begin your first minute by having team members say their names and briefly tell us the topic you chose to investigate (and why), and then use the remaining 4 minutes to tell us only the most important, interesting or surprising couple of results you found (because there may not be time to discuss everything you did), and what you might do differently the next time. It will be helpful to me (in considering the quality of your oral presentation for possible extra credit) if you hand me a handout of any visual aids you show (to save trees, put 4-6 slides per page).
4.) PROJECT WRITEUP should be 12-point Times New Roman double-spaced, checked for spelling/grammar. Aim for about 5-10 pages, not counting any appendix. Section headings should be clearly marked, assembled in the required order, and stapled (not put into a folder or sleeve, or attached by a paper clip or rolled-up corners) with an identification coverpage on top (that includes your full name, section, date, and a title that makes clear which option you chose and what your topic/theme was – in other words, not just “Stat 1380 final project”). The writeup needs the following sections in this order, each with its own clear heading:
· The first part after the coverpage is titled “Background” and describes the general question of interest and why it is important and what initial predictions or hunches, if any, you had before you collected data. Explain what population you chose (and why).
· “Method” describes what random sampling procedure was used (or why one was not or could not be used and what specific things you did to help make the sample as representative as possible of the population of interest).
· “Results” includes the objective findings (i.e., without interpretations – that goes in Discussion) of your survey, with appropriate numerical and graphical summaries included. (Graphs must be appropriately drawn, titled, and sized, with axes labeled. Also, some graphics like “exploding 3-D pie charts” are easily generated by Excel but can be visually misleading, so avoid those; a good article on pie chart pitfalls to avoid is http://onlinelibrary.wiley.com/doi/10.1111/j.1467-9639.2007.00306.x/pdf) There should be appropriate use of technology here. If some of the graphs or computer output really interferes with the flow of the paper, it is fine to place those in the Appendix and tell the reader in the Results section that this is where they can find them.
· “Discussion” includes interpretations, conjectures, or implications you make from your data; discuss any limitations of your data or method, discuss any difficulties or judgment calls you encountered and how you handled them, discuss what you’d do differently next time.
· The next part is titled “References” where you acknowledge (in APA form) help you obtained from any web or print sources or people to do this project. (If you had none, then omit this part.)
· The final part is the “Appendix” which can include a page or two of additional tabular, numerical or graphical summaries of your data, but should not include raw information that could identify anyone who was surveyed and should not include a large number of extra pages (do not attach 50 individual questionnaire sheets, for example). Include a copy of the actual survey instrument. Include a copy of any slides you used in your oral presentation (to save trees, put 6 slides per page). The final page of the Appendix must be the proposal form that I signed and returned before you collected data.
Your project must also stay within these 7 additional guidelines:
a) Data must be collected openly and recorded in such a way that respondents cannot be identified or linked to their answers if someone were to read your report or stumble upon your files.
b) All individuals must be informed. For example, if you do a survey, you could start with something like, “Good afternoon -- my name is _______ and I’m a student in Dr. Lesser’s Stat 1380 class at UTEP. Are you willing to take a 1-minute survey as part of my class project? The survey will not be published, your participation and responses will be confidential, and your name or other identifying information will not be recorded. Your participation is voluntary and you may stop at any time. Do you have any questions? Are you willing to participate?” (If you are considering a project that is not a survey, but more of an observational study or experiment, a more explicit written consent form may be necessary and I will assess that in your proposal form.)
c) In selecting people for your survey, you must avoid choosing people where there could be bias and/or vulnerability: co-workers, relatives, significant others, close friends, minors (people under the age of 18), prisoners, and individuals who cannot read/understand informed consent.
d) Your topic can be almost anything as long as it avoids issues which could reasonably be expected to be highly personal, controversial or sensitive, such as sexual behavior, drug usage, or underage drinking. But while you CANNOT ask about someone’s sensitive personal status or behavior (e.g., “How much do you weigh?”, “Do you carry a concealed handgun?”, “How much beer do you drink?”), you might be able to ask about the topic from a society or current events perspective (e.g., “Do you think obesity is a serious problem in El Paso?”, “Do you think the campus concealed carry law makes UTEP safer?”, “Do you think the drinking age in Juárez should be raised, lowered, or kept the same?”).
e) The proposal sheet (http://www.math.utep.edu/Faculty/lesser/1380proposal.html, or http://www.math.utep.edu/Faculty/lesser/1380proposal.pdf) must be approved by me before you collect any data or you may get no credit. This rule ensures that no one tries to do something with an inappropriate topic/design or unrealistic scope, as well as models the process that researchers use before they can begin a study. The sooner you submit your proposal form, the sooner I’ll give you ungraded feedback (usually at or by the very next class), and the more time you’ll have to do your project (or make adjustments to your proposal and resubmit it, if required). I recommend you start thinking about it now and turn in your proposal form to me in hardcopy on or before the deadline stated in the class syllabus (so that you can get my feedback and still have plenty of time to collect and analyze your data). Feel free to bounce more sophisticated ideas off of me, but the most common project here will be a survey of adults (be sure to ask a large enough sample, so that even if some refuse to participate, you’ll still have at least 50 people in the survey). Your proposal sheet will need to be turned in as the last page of your writeup.
f) There should be at least 2 categorical and at least 2 measurement variables on a topic of interest. Remember that saying how many men and women are in the sample is generally not 2 quantitative variables (“number of men”, “number of women”), but the results for 1 qualitative variable “gender” (for which “male” and “female” are two possible categorical values).
g) You should not only analyze each variable one at a time, but also explore any interactions or patterns among your variables, appropriately using a tool such as a two-way table, scatterplot, etc.
RUBRIC: these 10 bulletpoints are each worth 10 points (and up to 2 additional points can be earned due to the creativity/originality of your topic/design) for assessing your write-ups:
I. project presented orally (with each team member having a major non-silent role) and turned in in hardcopy form on or by due date [remember, the actual quality of the oral presentation is assessed only in a way that can help you if your final course average falls a point below a letter grade cutoff]; if you have a pre-approved reason not to be able to present in person on a presentation day, you need to arrange with me as soon as possible to present earlier (or, if necessary, upload as a YouTube video that we can show on the presentation day)
II. writeup includes all required sections (Background, Method, Results, Discussion, References, Appendix), with those headings clearly labelled and in the required order and does not unduly deviate from the target for number of pages; any references used are cited in complete APA style; quality of writing is basically acceptable (including not have spelling or grammar errors that are distracting or that make content unclear)
III. evidence (e.g., showing and discussing the actual procedure or instrument you used) that data was recorded in such a way that respondents cannot be identified or linked to their answers if someone were to read your report or stumble upon your files; Appendix of written report includes the actual data collection instrument (such as a blank questionnaire) or protocol, and proposal approval form that I signed; participants do not include people with major potential of bias or vulnerability; number of participants is appropriate for the type and requirements of the study
IV. The term random is used appropriately (and not inappropriately, such as “we picked 20 random students”) in the writeup and design of the study; also, the design should involve random selection (if you did a survey) or random assignment (if you did an experiment) if possible and should use and describe a thoughtful “next best thing” approach if it is not; also, if you did an experiment that was not double-blind, explain why you didn’t (or couldn’t) do double blinding and what steps you took to minimize threats to the validity of the experiment; if you did an observational study, be sure to explain why an experiment was not possible or ethical to do instead
V. If you do a survey, at least 2 categorical and at least 2 measurement variables are included and clearly defined (this includes NOT confusing a measurement variable with the count of a value of a categorical variable); the number of variables should not significantly exceed this, however, so that (1) there is not a major burden on participants, and (2) there is sufficient opportunity to do thorough exploration and analysis of the data that is collected; if you did an observational study or experiment, it may be reasonable to have one less measurement variable, but this would have to be approved in your proposal
VI. writeup not only analyzes each variable one at a time, but also explores whether there are interesting interactions or patterns between any 2 variables (for example, if you have 2 quant and 2 qual variables, there are 6 pairwise interactions to look at); for two qualitative variables, this could include a contingency table (e.g., p. 9), or conditional probabilities, or double bar graph (p. 183); for two quantitative variables, this could include a scatterplot or correlation coefficient; for one quant and one qual variable, this could include boxplots (or dotplots) that are side-by-side; or you may find a creative way to look at more than 2 variables at once, as we saw in http://www.gapminder.org/world/
VII. data collection procedure is appropriate and is described in sufficient detail for someone else to be able to do the collection again
VIII. appropriate graphical (i.e., at least one graph) and numerical summaries of the data are included in your written report and oral presentation, with labelling that is clear, self-contained (avoid “whiplash” where the reader has to keep moving her eyes back and forth), and not misleading or ambiguous; in addition to the graphics pitfalls listed in our textbook (e.g., p. 190), a good article on pie chart pitfalls to avoid is http://onlinelibrary.wiley.com/doi/10.1111/j.1467-9639.2007.00306.x/pdf; make sure that a graph does not use shadings of color or grayscale that are hard to distinguish on what you display and what you turn in; make sure that the summaries are not “a kitchen sink of every possible summary”, but are thoughtfully selected, sequenced, and discussed
IX. conclusions are clearly stated and appropriately qualified; you provide thoughtful reflections about the limitations (e.g., was there a high nonresponse rate?) or imperfections of data collection process, judgment calls that had to be made, things that might be done differently next time
X. assessment of your contribution to the group that your teammates provide to me confidentially