Increasing Partner Capacity
The Georgia Policy Labs believes government and school district staff are catalysts for creating evidence-based policies and programs that improve positive, equitable outcomes for children, students, and families. We draw on the expertise and experience of faculty and staff to facilitate capacity-building opportunities for our partners and the public policy community. Our goal is to provide trainings that respond to current needs and develop relevant skills that agencies can implement quickly. As part of our commitment to enhance evidence-based policymaking more broadly, we offer these trainings freely to the larger community.
R Project for Statistical Computing
Introduction to R and tidy
In this workshop, you will be introduced to the tidyverse, a set of R packages that work together to make loading, cleaning, manipulating, reshaping, visualizing, and analyzing data fun and intuitive. You will learn how to import data from CSV files, Excel, Stata, SPSS, and SAS with the readr, readxl, and haven packages; filter, summarize, reshape, rearrange, and manipulate data with the dplyr and tidyr packages; create beautiful graphs and visualizations with the ggplot2 package; and write reproducible HTML, Word, and PDF reports with R Markdown and the knitr package.
In this workshop, you will learn about the grammar of graphics–a special language for describing data-based graphics–and how to use the ggplot2 package in R to create beautiful and powerful data visualizations. You will learn how to map columns in a data set to specific elements (or aesthetics) in a graphic; learn about the key elements of the grammar of graphics and how to translate the grammar to R code; learn best practices for creating truthful, accessible, and well-designed data visualizations; create interactive HTML graphics with the plotly package; create animated graphics with the gganimate package; and create simple HTML dashboards with the flexdashboard package.
Program Evaluation and Causal Inference
In this workshop, you will learn about exciting new statistical methods that allow you to make causal inferences from observational data: causal diagrams (or directed acyclic graphs) and do-calculus. You will then apply this language of causal inference to program evaluation and learn how to measure the effectiveness and outcomes of social programs with several different statistical methods. You will learn how to create causal models with directed acyclic graphs (DAGs); use basic do-calculus to close and isolate backdoor paths between program treatment and program outcome variables; apply the language of causal models to core econometric techniques, including multiple regression, randomized control trials (RCTs), difference-in-differences (diff-in-diff, or DiD), and regression discontinuity (RD) research designs; analyze results from a randomized control trials; analyze results from an observational study using do-calculus and regression; analyze results from an observational study using difference-in-differences; and analyze results from an observational study using regression discontinuity.
Introduction to Python for
In this workshop, you will learn the basics of Python, including getting started with Anaconda and JupyterLab; writing, executing, and annotating code; and using Python syntax, objects, and data types. This workshop also focuses on learning more about data analysis using the Pandas module to load and import data, view and subset data, analyze descriptive statistics, and run crosstabs and chi-squared tests.
Python for Statistical Modeling and Plotting Data
In this workshop, you will learn more about statistical modeling, including the general logic behind fitting models in Python, how to use the Statsmodels module to generate linear models, and how to use the Scikit-Learn module to generate models. Additionally, you will learn more about data visualization using the Seaborn module, focusing on an overview of plotting syntax using the Matplotlib and Seaborn modules, creating simple univariate and bivariate plots, adjusting plot aesthetics, and layering multiple plots together.
Python for Data Wrangling, Text,
In this workshop, you will learn more about data wrangling, including merging data sets and transforming variables (i.e., recoding, relabeling, and binning). You will also learn new skills like installing new modules and libraries (e.g., Plotly) and displaying code outputs on Github. You will also work with text data and explore string detection and term frequencies.