Project Assignment

For the project you will create, in small groups, a thorough analysis of a particular dataset. For each dataset, the goal is to tell an interesting story about the dataset. The analysis project will have components completed by the group and by each individual student. Each group will prepare a formal written report and will present their project to the class in a 15-20 minute presentation. In this presentation, each group member will have 2-3 minutes to present his/her topic. Additionally, each group will create a 1 page handout (two-sided, if desired) to accompany the group presentation. The handout provides an opportunity to summarize key findings, show central figures, and/or provide greater detail or explanation about pieces of the project.

Each group will be expected to hand in the following items

1. a PDF file containing the data analysis write-up that has been typeset using RMarkdown/knitr,
2. a single paper copy of the final report,
3. a PDF file containing the presentation handout, and
4. a single paper copy of the presentation handout.

The instructor will pick teams and assign each team a particular dataset.

Guidelines for the project write-up

As a team, the group should read at least 1-2 papers that describe the dataset, to verse yourselves in the background of the dataset. The group will then assemble an outline of the key elements of the story that you want to tell about the dataset and which team member will be responsible for at least one element. The general idea is that each element focuses on one key observation or insight about the dataset. Don’t overburden your team with a story with lots of different elements. Try to tell a short, compelling story with a small number of elements. Ideally, each each teammate has one and only one. The elements should complement each other and together tell a coherent story about your dataset. At least one element (and preferably more) must include a regression model.

Overall, the project write-up should be written in clear, concise prose, suitable for publication in a scientific journal. No code should be shown in the write-up, although it is expected that results provided in the write-up will be dynamically generated (i.e. if you report the results from a regression, you are calling the results directly from R and not inputting the numbers directly into the .Rmd file). You will need to be very judicious in your choices of what to include in the write-up, only leaving the items most central to the write-up’s overall goals and storyline.

Please follow the structure and page limits given below:

• cover page (title, names, table of contents): 1 page
• group data analysis (including tables/figures): 2 pages
• each individual data analysis (including tables/figures): 2 pages
• conclusion/discussion: 1 page

No page-cramming: i.e. nothing smaller than 11-point font, no less than 1 inch margins all around.

Group Data Analysis

The group-written data analysis will provide a brief summary of key features of the dataset. You should define and summarize each variable that will be used, either in a table or graphically, or both. Any central hypotheses or relationships that will be tested or explored should be defined here. At least a few sentences of context and description of the dataset should be included. This section should include a few tables and/or figures and should be no more than two pages long. Description of how missing data was handled (if you have any) should be included here (see more detail below).

Individual Data analyses

Each member of the group will lead a particular thread of analysis and/or dataset exploration that helps tell a story about the assigned dataset. The write-up for each individual data analysis should not exceed two pages, including tables and figures. The individual write-ups should be focused on a key element of, relationship within, or insight about the dataset.

Examples of elements are:

1. identifying and describing, using appropriate analytical tools (e.g. a regression model) to descibe relationships important variables in your dataset. This might include fitting a regression model and interpreting the output, with particular attention paid to a relationship that you know scientifically to be a question of interest.
2. showing, using an in-depth or sequence of data visualizations, an important feature of relationship within your dataset.
3. creating and evaluating predictions from your model.
4. fitting and describing key features of this a regression model, such as discussing the possible influence of outlying points, missing data, non-linear relationships, etc…

The data analysis write-up will contain a section for each group member’s analysis and a conclusion that summarizes the results.

Project grading

Your project grade makes up 40% of your final grade for the class and will be calculated as follows:

• The final product produced by the group 50 points
  • 20 points: group data summary (clarity of data summary, quality of graphics/tables, adequate and accurate explanations of data, specific hypotheses defined, results summarized in conclusion/discussion section)
  • 15 points: group presentation (time limits adhered to, project clearly summarized and defined, key findings highlighted, polished presentation, handout provides useful detail and/or explanations)
  • 10 points: uniformity of presentation (individual write-ups have same look and feel, the project feels like a single work, not too disjointed, topic coordination a plus)
  • 5 points: project details (page limits adhered to, appropriate sectioning, etc…)
• Individually prepared data analysis 50 points
  • 35 points: overall quality (clear and accurate description of methods/models used, correct implementation and interpretation of method(s) used, appropriate use of equations to show what methods/models have been used, appropriate use of graphics/tables to support central results, summary of key results)
  • 15 points: individual presentation (clear statement/summary of goals and central results, use of figures rather than text to illustrate central ideas, time limit adhered to)

To evaluate group participation and contributions, I will be using the following approach to evaluate each of your contributions to the project. Each student will be given 100 points to allocate among your teammates (excluding yourself). The more points you give to a teammate, the more you are indicating they contributed to the project. You cannot allocate the same number of points for any two team members. I reserve the right to intervene to correct gross imbalances in allocations if necessary. The number of points that you receive from your teammates will be summed, divided by 100, and then used as a multiplier on the final grade for the group project.

As an example: Your group receives 40/50 points for the “final product produced by the group”. You have three teammates who give you scores of 35, 40 and 30, respectively. Therefore, you receive a total of 105 points from your teammates. So your final “group” grade is (40/50) * (105/100) = 0.84 = 42/50.

Deadlines

• Wed Apr 6: Group ouline and individual topics proposed (1 paragraph summary, submitted to instructor on Google Drive)
• Wed Apr 13: Draft of group data description write-up due, 5pm
• Mon Apr 18: Draft of individual data analysis due, 5pm
• Thu Apr 21: Group 1 (and Group 2?) present, project handed in on Wed 4/27 by 5pm
• Tue Apr 26: Group 3 (and Group 2?) presents, project handed in on Wed 4/27 by 5pm