Final Meetup

class: center, middle, inverse, title-slide

# Final Meetup
## DATA 606 - Statistics & Probability for Data Analytics
### Jason Bryer, Ph.D.
### May 7, 2025

---
# Final Exam

* Is now available on Brightspace.

* Due by midnight May 11th.

* You may use your book and course materials.

* We expect you to complete the exam on your own (i.e. do not discuss with classmates, colleagues, significant others, ChatGPT, etc.)

* There are two parts:
	1. Part one multiple choice questions and short answer questions.
	2. Part two has a small data set to analyze with R, then answer some interpretation questions.

* Put your answers in the Rmarkdown file and submit the PDF file. **Please do not post your answers online!**

---
# Announcements

* You should join the New York Open Statistical Programming Meetup group. [https://nyhackr.org](https://nyhackr.org)
  * They meet monthly, usually at NYU. [George Hagstrom](mailto:George.Hagstrom@cuny.edu) has been organizing a group of MSDS students meeting up there each month.
  * Their annual conference will be August 26th and 27th: [https://nyhackr.org/events.html](https://nyhackr.org/events.html)

* The Joint Statistical Meeting will be August 2nd to 7th in Nashville: [https://ww2.amstat.org/meetings/jsm/2025/](https://ww2.amstat.org/meetings/jsm/2025/)

* useR! will be virtual August 1st and in person (Duke University) August 8th to 10th: [https://user2025.r-project.org](https://user2025.r-project.org)

---
# Propensity Score Analysis <img src="images/hex/psa.png" class="title-hex">

My statistical research interest is in propensity score methods. Propensity score analysis (PSA) is a quasi-experimental design used to estimate causality from observational studies.

Here are some resources for PSA:

* PSA [Github repository](https://github.com/jbryer/psa) includes slides slides and Shiny application: https://github.com/jbryer/psa
* Early version of an [Intro to PSA](https://psa.bryer.org) book: https://psa.bryer.org
* Recording of a talk given in Fall 2023 for the NYC Meetup group here: https://www.youtube.com/watch?v=JLV4mtFhRMM

.pull-left[
<img src='images/hex/psa.png' height='100' align='left' style='padding-right:20px'> `multilevelPSA`<br/> [Multilevel PSA]((http://jason.bryer.org/multilevelPSA)
<br/><br/>
<img src='images/hex/TriMatch.png' height='100' align='left' style='padding-right:20px'> `TriMatch`<br/> [Matching with non-binary treatments](http://jason.bryer.org/TriMatch)

]
.pull-right[
<img src='images/hex/PSAboot.png' height='100' align='left' style='padding-right:20px'> `PSAboot`<br/> [Bootstrapping PSA](http://jason.bryer.org/PSAboot)
<br/><br/>
<img src='images/hex/PSAgraphics.png' height='100' align='left' style='padding-right:20px'> `PSAgraphics`<br/> [Graphical analysis of PSA](http://jason.bryer.org/PSAgraphics)
]

---
# R Packages

Here is list of some other R related projects I have worked on:

<img src='images/hex/likert.png' height='70' align='center'> [`likert`](https://github.com/jbryer/likert) - Analysis and Visualization of Likert Based Items  
<img src='images/hex/ShinyQDA.png' height='70' align='center'> [`ShinyQDA`](https://github.com/jbryer/ShinyQDA) - R Package and Shiny Application for the Analysis of Qualitative Data  
<img src='images/hex/medley.png' height='70' align='center'> [`clav`](https://github.com/jbryer/medley) - Predictive modeling with missing data  
<img src='images/hex/clav.png' height='70' align='center'> [`clav`](https://github.com/jbryer/clav) - Cluster Analysis Validation  
<img src='images/hex/IRRsim.png' height='70' align='center'> [`IRRsim`](https://github.com/jbryer/IRRsim) - An R Package for Simulating Inter-Rater Reliability  
<img src='images/hex/mldash.png' height='70' align='center'> [`mldash`](https://github.com/jbryer/mldash) - Machine Learning Dashboard  
<img src='images/hex/FutureMapping.png' height='70' align='center'> [AmplifyApp](https://amplifyapp.org/en), [dashboard](https://amplifyapp.org/en), and [Future Mapping NYC](https://futuremapping.org)

---
# DAACS

[The Diagnostic Assessment and Achievement of College Skills](https://daacs.net) (DAACS) is a suite of technological and social supports to optimize student learning. DAACS  provides personalized feedback about students’ strengths and weaknesses in  terms of key academic and self-regulated learning skills, linking them to the resources to help them be successful students. This is currently supported by a five-year $3.8 million grant received in 2021 from the Institute of Education Sciences to test the efficacy at three institutions.

Applications of Data Science:

* We use natural language processing and predictive models to machine score the essays.
  * We had a student this semester work with us to explore whether we can detect AI generated essays specific to the DAACS writing prompt (answer: we)
* We use DAACS data to estimate "risk scores" for students failing so we can target them with resources to help them be successful.
  * Related to this, we have developed a new R package for estimated predictive models with missing data, see [medley](https://github.com/jbryer/medley).

---
# Thank You

This has been a great semester. Please don't hesitate to reach out:

<br/>

You can download all course materials on [Github](https://github.com/jbryer/DATA606-2025-Spring). Click the [clone or download](https://github.com/jbryer/DATA606-2025-Spring/archive/master.zip) link to download a zip file.