Final project

Tasks

by 15 March you should:

decide which data set/which questions you are going to pursue
pick a short name for your project (letters, numbers, and underscores only please) and create a new repository with this name; whoever creates it should invite the other project participant(s) as well as @bbolker and @dushoff.
upload your data (if possible)
write a README.md file for your repository that says briefly what you plan to do and says where the data is/are

by 22 March you should:

update your README with thoughts and questions about next steps
Start a file called journal.md where you record anything you do that involves looking at patterns in the data (i.e., anything but data cleaning). The journal is also a good place to record plans for how you will proceed with your analysis
- (for example, “if the diagnostic plots look OK, we will look at the summary table and use the clarity of the response to phosphorus as our main outcome”. (See Simmons, Nelson, and Simonsohn (2011).)

Presentations

We will have presentations on Thu 4 April. Presentation slots will be assigned at random. Schedule will be posted Wed morning. Let us know well in advance if you have constraints or problems.

You should plan to talk for 10-12 minutes and take questions for 3-5 minutes.

The presentation is not expected to be a finished work; you will still have three more weeks to work on your writeup. Outline your scientific question, the approaches you are taking, and outstanding questions about how to go forward. You can include one or two results (or preliminary results), if you have them.

Please plan to finish on time. Time is short, and we want to be able to comment when appropriate.

The presentation is worth two ordinary assignments.

Writeup

In the final project writeup, you should introduce a scientific question and present the results of a group of statistical analyses. that answer the question. You don’t need to write a formal paper, but you should present both the scientific and statistical background clearly. Discuss what choices you made, and if necessary what further statistical steps you should take in the future.

Your writeup should be posted to your repository as a PDF file. Please do not include R code in your writeup. All analyses should be clearly documented and reproducible; your README.md file should describe the location of code and data files, and the steps necessary to reproduce the analyses.

The writeup is due at noon on Thu 25 April. It is worth four ordinary assignments.

Project proposals

Please write up a short (1-2 paragraph) proposal for a group project in biological data analysis and post it to the Project descriptions Teams channel by Mon 4 Mar.

Your project should focus on the biological questions you want to answer, and the data that you have to answer them with; it can say a little about the quantitative/statistical techniques you will use, if you already have ideas in that direction.

If you have data to use, it must be shareable at least with the instructors and with other students working on your project (and you must be able to share the results of your analysis with the whole class)
If you don’t have access to any data (yet), or your data are sensitive/private, you can:
- try to get similar data from a colleague, or from a published analysis
- structure your problem around a simulation analysis, i.e. construct randomized ‘data’ that will help you understand whether/how a particular set of statistical techniques actually works when analyzing data with a given effect size and noise level

The more interesting and engaging your writeup is, the more likely that someone else will want to work on your data with you. If you do not want to work on the question that you proposed, please state that clearly in your proposal.

This is supposed to be an initial set of ideas; you’re not required to do exactly what you describe here (e.g. if it turns out to be infeasible, or something else ends up being more interesting).

Feel free to reach out to the instructors if you want to clarify the parameters of the assignment.

We will help organize groups based on interests and make recommendations about which of the proposals from that group are the most feasible and appropriate for this course.

Simmons, Joseph P., Leif D. Nelson, and Uri Simonsohn. 2011. “False-Positive Psychology Undisclosed Flexibility in Data Collection and Analysis Allows Presenting Anything as Significant.” Psychological Science 22 (11): 1359–66. https://doi.org/10.1177/0956797611417632.