10-projects
Case Studies & Final Projects
Q&A
Coming Soon
Course Announcements
Coming Soon
Agenda
- Case Studies
- Final Project
Case Studies
Biomarkers of Recent Use
- We’ll use data from this paper:
Hubbard et al. Biomarkers of Recent Cannabis Use in Blood, Oral Fluid and Breath. Journal of Analytical Toxicology. 2021. Link to paper.
OpenCaseStudies
- OpenCaseStudies
- Uses R/the
tidyverse
- asks public health-centric questions
- goal: to teach statistical analysis/data science through case studies
What We’ll Do
For each case study (2), during lecture:
- Stats: (1-2d)
- Background, Data & Wrangling (1-2d)
- EDA & Analysis (1-2d)
. . .
For each case study:
- you’ll also work with case study data in lab.
- you’ll work in assigned groups of ~3 students to complete a data science report
. . .
I will share previous student examples and we’ll discuss pros and cons in coming lectures.
Data Science Reports
With your group, you will:
- carry out all steps of the analysis
- some code will be taken directly from lecture
- add text/organize into a report
. . .
- have to extend the case study
. . .
This should be written at the level of a data science-knowledgeable undergrad.
General Communication Submission
This is (intentionally) very open-ended.
You need to communicate the most important aspect/finding/part(s) of your case study to a general audience (any undergrad).
. . .
What might this look like?
- short TikTok like video
- brief Youtube video
- slides for an Instagram post
- X (Twitter) thread
- poster to be displayed next to an elevator
- poster to be put on public bulletin boards
- effective email communication
What does extend the case study mean?
You’ll need to do something more on the topic beyond what is presented in class.
. . .
Examples:
- Asking an additional question and answering it from the data provided
- Finding an additional dataset and using it to add to the case study
- Generating a handful of additional and very informative visualizations (beyond what’s presented in class)
Grading
Graded on:
- content (code, text, viz)
- report: effective written communication (clarity/content > grammar/spelling)
- extension carried out
- effective general communication (effectively conveys message to a general audience)
Final Project
Final Project Logistics
- will be completed in groups of 3-4 students
- you get to choose the group
- I will ask Monday week 7 for your final project groups (If you are not in one, I will help)
- You will submit a proposal week 8.
- Final projects are due during Finals week
What is the final project proposal?
- A short Google Form
- you’ll submit your topic and a few details about that topic (depending upon which option you choose)
- Your idea can change after you submit your proposal
- This has been added to help you start your project before finals week.
Final Project Details
Two possible paths:
- Create a technical presentation on a statistics topic and/or an R package.
- Carry out a data analysis
Option 1: Technical Presentation
- .Rmd document used to make slides
- “Teaches” the details of the R package/statistics topic
- Demonstrates how to use the package and/or carry out the statistical analysis in R
- Topic/Package must go beyond what was taught in this course or what you should have learned in an intro stats course
- Presentation Length: 10-15min
Option 2: Data Analysis
- .Rmd document used for data science report
- Asks a question, finds data, analyzes data (basically: a mini case report, but you find the data and formulate the question)
- Presentation Length: 3-5min (brief summary of the full report)
Where/when for this presentation?
- Submit by Tues of finals week at 11:59 PM
Should I be working on my final project now?
…probably not
. . .
But, you should start thinking about/getting a group of 3-4 people together. You’ll need to submit who’s in your final project group Monday of week 7.
. . .
You’ll need to have a general plan for your final project around wk 8. You’ll submit a “proposal” Monday of week 8.
What is the “general audience” communication?
Consider who the audience would be -> design for them
. . .
For example, if you present on an R package, who would benefit from knowing about this package? How would you reach them? What can you design to inform them of what it is and get them to use it?
. . .
Or if you do a data analysis on a particular topic, what would you want others to know? How would you communicate that?