Program Content

The program will provide a hands-on introduction to data analytics topics ranging from basic SQL and Python coding to building and interpreting machine learning models to tackle policy problems. During the program, we will use a mix of confidential datasets in the course materials.

The training will take place at the main College Park, MD, campus of the University of Maryland. 

Table of contents:

• Communications

• Pre-course material

• Data documentation

• TANF readings and references

• Training program

  • Site logistics

  • Module 1: Oct 31 - Nov 2 & 5-6

  • Module 2: Dec 5-7 & 10-11

• Projects - presentations in-person Dec 11, final write-up due Friday, Dec 21

COMMUNICATIONS

We primarily use three media to distribute information and communicate amongst the group: this website, e-mail and Slack. In general, the instructor team will respond quickly to either e-mail or Slack messages; however, we tend to prefer Slack for technical issues and sharing snippets of information.

E-mail addresses

• Group email account

  • dataanalytics@coleridgeinitiative.org - the group email for program instructors and coordinators; general program, logistics and ADRF system questions should be directed here (ADRF is our computing platform, see the computing environment page for more info)

  • support@applieddataanalyticsprogram.org - ADRF support account from which system communications will be sent

• Instructor email addresses:

  • Rayid Ghani - rayid@uchicago.edu

  • Clayton Hunter - clayton.hunter@nyu.edu or clayton@coleridgeinitiative.org

  • Brian Kim - kimbrian@umd.edu

  • Frauke Kreuter - fkreuter@umd.edu

  • Avishek Kumar - avishekkumar@uchicago.edu

  • Julia Lane - julia.lane@nyu.edu

Slack

We use Slack extensively and it is often the best way to get in touch with us, the team is coleridge-initiative.slack.com and by now you should all have received invitations to join (if you have not please let us know). If you are unfamiliar with Slack, it has different "channels" to help organize conversations; previous classes have used Slack in various ways (eg for a channel for Python specific questions), but the two primary channels we expect to use are (i) the "class-6-uchicago" channel for general class discussion and sharing documents and (ii) the "adrf-tech-support" channel for any technical support in accessing the ADRF.

PRE-COURSE MATERIAL

Our collaborators at the University of Maryland and the University of Mannheim have prepared some introductory material for you to work through prior to the program. The material is presented in a four-week, short course format; two weeks for SQL and two for Python. We expect the material for each week to take a maximum of 4 hours total to work through. 

DATA DOCUMENTATION

• Illinois Department of Employment Services (IDES)

  • Documentation link

• Illinois Department of Corrections (IDOC) admissions & exits: link to pdf

  • More details on EDUCLVL and HCLASS fields:

    • link to pdf of notes

    • link to Excel of Offense Codes

  • Primary assignment codes

  • Birthplace codes

• Illinois Department of Human Services (IDHS): documentation link

• Missouri State LEHD Wage and Employers Data

• Census LEHD Origin-Destination Employment Statistics (LODES)

TANF Readings and resources

• Leaving Reach Up: How did the experiences of Vermont’s 2013 welfare leavers compare to earlier leavers?

• Colorado Works Exit Survey Project Summary of Fiscal Year 2016/2017

• Life After Welfare: 2017 Annual Update (Maryland)

• Mandatory Work Requirements for Recipients of the Food Assistance and Cash Assistance Programs (Florida, Jan 2018)

• Family Employment Program (FEP) Redesign Study of Utah 2014: Final Report

• WorkFirst Wage Progression and Returns Report: through second quarter 2012 (WA)

• WorkFirst Wage Progression and Returns Report: through first-quarter 2016 (WA)

• REPORT TO THE LEGISLATURE WorkFirst Wage Progression and Returns Report: Through fourth-quarter 2016 (WA)

• OPRE’s self-sufficiency research clearinghouse

• Illinois TANF program - State website

• HHS Luminaries series of videos: link

Training program

Site Logistics

Training location: see agenda below

Lefrak Hall address: 6903 Preinkert Dr, College Park, MD 20740

Links to additional information

1. The UMD Visitor Parking Map

2. Mobile Parking App Information

3. How to use the Pay Stations

4. Campus Shuttle Bus information

5. Washington, DC Metro map

Team list

WEEK 1: Oct 31 - Nov 2 & 5-6 (9am - 4pm eastern)

Wednesday-Friday: 1208 Lefrak Hall, University of Maryland

Oct 31 - Program introduction and overview

• 09:00 - 09:30 Welcome and Introductions (morning slides)

• 09:30 - 09:45 Goals for the program

• 09:45 - 10:45 Project scoping

• 10:45 - 11:00 Break

• 11:00 - 12:00 Introduction to ADRF and Security Training

• 12:00 - 01:00 Lunch

• 01:00 - 01:15 Review of the morning (afternoon slides)

• 01:15 - 03:30 Buffet of Analytics topics

  • Topics we will cover

  • Topics we will NOT cover

• 03:30 - 04:00 Logging in to the ADRF

Nov 1 - Data exploration & Visualization

• 09:00 - 09:15 Welcome and overview of the day

• 09:15 - 10:30 Databases (lecture slides, normalization slides)

• 10:30 - 10:45 Break

• 10:45 - 12:00 Data Visualization (lecture)

• 12:00 - 01:00 Lunch

• 1:25 - 04:00 Hands-on data exploration with SQL & Python

Nov 2 - Record Linkage

• 09:00 - 10:15(?) Data Visualization (notebook)

• break

• 10:30(?) - 12:00 Record Linkage (lecture, McDonald “An Introduction to Probabilistic Linkage”)

• 12:00 - 01:00 Lunch

• 01:00 - 01:45 Guest Lecture, Rick Hendra: Research Evidence on TANF and Employment (slides)

• 01:45 - 02:45 Data preparation: creating "labels"

• 02:45 - 03:00 Break

• 03:00 - 04:00 Data preparation: creating "features"

Monday-Tuesday: 2208 Lefrak Hall, University of Maryland

Nov 5 - Introduction to Machine Learning

• 09:00 - 12:00 All (almost) of Machine Learning (slides)

• 12:00 - 01:00 Lunch

• 01:00 - 02:45 Walk through of example ML prediction model

• 02:45 - 03:00 Break

• 03:00 - 04:00 Project discussion: label (outcome) definition and features

Nov 6 - Text Analysis

• 09:00 - 10:30 Text Analysis (lecture)

• 10:30 - 10:45 Break

• 10:45 - 12:00 Text Analysis (notebook)

• 12:00 - 01:00 Lunch

• 01:00 - 01:15 Project goals and timeline (slides)

• 01:15 - 04:00 Project work (exploration, discussion, planning)

WEEK 2: Dec 5-7 & 10-11 (9am - 4pm eastern)

Wednesday-Friday: 1208 Lefrak Hall, University of Maryland

Dec 5 - Machine Learning - methods

• 09:00 - 12:00 Machine Learning lecture - recap (slides) and methods (slides)

• 12:00 - 01:00 Lunch

• 01:00 - 02:00 Project discussion - model set-up

• 02:00 - 04:00 Project work

Dec 6 - Machine Learning in Practice

• 09:00 - 12:00 Machine Learning lecture - model selection, evaluation (slides), and bias (slides)

• 12:00 - 01:00 Lunch

• 01:00 - 04:00 Project work

Dec 7 - Privacy and Confidentiality

• 09:00 - 10:30 Privacy & Confidentiality lecture (slides)

• 10:30 - 10:45 Break

• 10:45 - 12:00 Disclosure review & export requests

• 12:00 - 01:00 Lunch

• 01:00 - 04:00 Project work

Monday-Tuesday: 2208 Lefrak Hall, University of Maryland

Dec 10 - Inference

• 09:00 - 10:30 Inference lecture (slides)

• 10:30 - 10:45 Break

• 10:45 - 12:00 Project work

• 12:00 - 01:00 Lunch

• 01:00 - 04:00 Project work

Dec 11 - Final Presentations

• 09:00 - 09:10 Program recap

• 09:10 - 11:00 Finalize project presentations

• 12:00 - 01:00 Lunch

• 01:00 - 04:00 Final Presentations

Projects

Final project presentations will be held in person on December 11. Each team will have 20 minutes to present followed by 10 minutes of Q&A.

Export requests should be submitted by end of day Monday, December 17.

Final project reports are due Friday, December 21.