STA490 2020-21 Expectations for Collaborative Project— STA490Y 2020–21 This collaborative project is an important component of STA490. Across both sections of the course, students will be working on one of several projects. You will be assigned to a small group of 3–4 students who will work on the same project and be supported by a PhD student project mentor. Most of the time, you and your group will meet with your project mentor to discuss challenges you’ve encountered and come up with strategies for addressing these, but from time-to-time, you’ll meet with your project mentor one-on-one to discuss your project or meet with the project collaborator (along with all other student teams working on the same project). There are several major components to this project: A. Project Logs B. Exploratory Data Analysis C. Statistical Analysis D. Written Report E. Presentation While discussion with your group members in project meetings is required and will count towards your course participation score (and discussion outside these meetings is encouraged as well!), you will make your own analysis decisions and submit your work individually, apart from the presentation. The project presentation will be a group presentation, which you will do with the 3–4 members of your project group. Project meetings As you work on your projects over the next several months, you will have project meetings / one-on-one consultations with your project mentors most weeks. Your project meetings will always be scheduled on the day/time corresponding to your section but will generally be shorter than 2 hours to allow project mentors to spend time with each of their three groups. - L0101: Tuesdays 10:10–noon ET - L0201: Mondays 5:10–7:00 p.m. ET For more details about the dates/times of your project meetings, refer to the Weekly materials pages for Fall and Winter. A – Project logs For your STA490 collaborative project, the project log is a diary of the work you do related to the project. The purpose of the project log is: (1) to document your work; (2) to keep track of any questions you have, as well as the answers (when you get them); and STA490 2020-21 (3) to help you to organize your thoughts before and after each meeting. Each of you will have access to a private discussion forum on Quercus, where you will post your logs as you work on your project over the course of the year (your classmates will not be able to view your log, but your project mentor will). Once you begin working on your project, you will, at minimum, post a pre- meeting log no later than 3 hours before your project meeting or one-on-one meeting and a post-meeting log by no later than 12 hours after your meeting. BUT we would recommend you take 10–15 minutes directly after your meeting to get your action points down, that should be all the time you need to spend on post-meeting logs. You are encouraged to post additional entries in your log between meetings to keep track of your work, but this is not required. In your pre-meeting log, you should include the following information: - A bullet point list of what you have done since the last meeting - Questions or challenges you encountered, which you’d like to discuss in your next project meeting (or one-on-one meeting) - Questions you’d like to ask the collaborator, either during the meeting (if it is a collaborator meeting) or potentially via email, after discussing with your fellow group members and project mentor - Solutions to problems you encountered either that week, or from a previous week (including references to textbooks / websites / articles / R code if appropriate) In your post-meeting log, you should include the following information: - Important information you learned during the meeting - Questions that came up during the meeting, which you or a group member will look into and report on at the next meeting - Ideas for next steps In short, the purpose of the pre-meeting log is to organize your thoughts to get the most out of your meeting, and the purpose of the post-meeting log is to generate a list of action items for you to work on for your project before the next meeting. B - Exploratory Data Analysis (EDA) Once you obtain the data, you will carry out exploratory data analysis (EDA). There will be two phases to this assessment. A peer review of your draft EDA followed by a final submission to your project mentor. Draft EDA peer review Create Phase (pdf submission) 11:59 a.m. ET Friday, October 16 Submit EDA report as a pdf through Quercus (using the peerScholar interface) Assess Phase (Peer Feedback) 11:59 a.m. ET Friday, October 23 Submit through Quercus (using the peerScholar interface) Note: You will only be receiving marks for i) completion and ii) quality of the feedback you give. We will not be marking your EDA draft and you can continue working on it up until the final submission, but the more you have done to submit for peer review, the more useful your peer’s feedback will be to you. You STA490 2020-21 will be assigned to review 2 EDAs from a DIFFERENT project and asked to provide ratings as well as short feedback. Final EDA submission to project mentor You will submit an .Rmd document (not a pdf as with the previous phase) and the data you used in to carry out the analysis. As you work on your EDA, be sure to do the following: - Document all data cleaning decisions you make. - Summarize the main reason(s) for analyzing these data (in your own words). What are the goals of the analysis? - Summarize the main features and patterns in the data. For all plots and summary tables that you include, you should write a description of the main features observed, in the context of the problem. - Identify any possible issues that you observe in the EDA that will affect the statistical analysis that you will do next. Your .Rmd document should be organized clearly such that the reader can easily understand why you produced each plot/table (or series of plots/tables), and what insights you learned from each plot/table. You should write a few sentences after each plot/table commenting on what you observe, specifically focusing on aspects which may inform further exploration of the data (e.g. unusual values, potential associations worthy of further investigation, etc.) and aspects which may have an impact on statistical analyses (e.g. missing values, unusual values). Evaluation of your EDA will be based on organization, your effectiveness at communicating relevant insights, and the reproducibility of your work. Note that you are submitting your .Rmd and your data but not the rendered pdf or html document. Your project mentor will be knitting your .Rmd to create their own copy of your output to review. They must be able to knit your file without errors in order to mark your work. C – Statistical Analysis You will analyze the data using appropriate statistical methods to answer the research question(s). Be sure to verify model assumptions and fit, interpret your results in the context of the research question(s), and consider limitations of the statistical methods you employ. You will conduct your analysis in R and prepare a .Rmd file containing your R code (clearly documented), and brief written commentary motivating your analyses and what you observe. Evaluation of your statistical analysis report will be based on the appropriateness of the methods you employ to answer the research question(s), on your written justifications for choosing these methods (acknowledging limitations), and your interpretation of the results. You will submit the .Rmd document through Quercus, and your project mentor will knit your .Rmd document to create their own copy of your output to review. D – Written Report Your written report will be addressed to the collaborator who posed the research question (not a statistician). There are no strict rules for how long your report should be, but we expect that it will be about 5–7 pages long (including figures and tables). Note that you will submit a complete draft (including title page, abstract, introduction, methods, results, and conclusion/discussion) a few weeks before your STA490 2020-21 final report is due for feedback both from classmate(s) working on a different project and from your project mentor. You will have time to incorporate this feedback before submitting the final version of your report. E – Presentations There will be three class presentations related to the project—one third of the students working on each project will be assigned to each presentation (more information to come later). The first presentation (November 5th) will focus on introducing the project, describing the research question(s), and discussing potential challenges and strategies or addressing them. The second presentation (December 3rd) will focus on describing statistical methods under consideration to address the research question(s), and the third presentation (February 5th) will focus on explaining the final analysis method selected and on interpreting the results. You will submit slides on Quercus by 11:59 p.m. ET the day before your presentation Key dates and deliverables Project logs Various dates - Pre-meeting log due 3 hours before each project meeting - Post-meeting log due 12 hours after each project meeting Post on your personal project log discussion board (only you, your PhD mentor, and instructor will have access) 10 pts EDA (draft) Create phase due 11:59 a.m. ET Friday, October 16 Assess phase due 11:59 a.m. ET Friday, October 23 Submit EDA report as a pdf through Quercus (using the peerScholar interface) Submit feedback through peerScholar 2 pts EDA (final) October 30, 11:59 p.m. ET Submit .Rmd & data file(s) on Quercus 13 pts Statistical analysis February 5, 11:59 p.m. ET Submit .Rmd & data file(s) on Quercus 20 pts Written report Draft Report #1 (submission & feedback on peers’ draft reports) February 26, 11:59 p.m. ET (your draft) March 5, 11:59 p.m. ET (feedback on peers’ draft reports) Submit .pdf, .doc, or .docx file on Quercus Submit feedback through PeerScholar 5 pts Draft Report #2 March 12, 11:59 p.m. ET Submit .pdf, .doc, or .docx file on Quercus 10 pts Final Report March 31, 11:59 p.m. ET Submit .pdf, .doc, or .docx file on Quercus 30 pts Project Presentation (as assigned) Presentation #1 Presentation #2 Presentation #3 November 5 (in class) December 3 (in class) February 5 (in class) Submit slides on Quercus by 11:59 p.m. ET the day before your presentation 10 pts
欢迎咨询51作业君
