This repository explores a variety of analytical methods applied by MSI staff. Follow a link below to explore analytical methods, or a novel dataset, or both.
Exploring ACLED conflict events
Exploring ACLED conflict events in South Sudan
Working with strings to cheat at help with Wordle
Line graphs in R
Tutorial content on crafting high-quality line graphs in R
Aessing the new class of d-i-d estimators
Turns out we've doing a lot of our regressions wrong. The common program evaluation setup of a two-period difference-in-differences generalizes across multiple groups and time periods to the two-way fixed effects estimator, but this estimator has only recently been shown to produce biased estimates in settings of staggered adoption of treatment and/or heterogeneous treatment effects. A new class of d-i-d estimators to the rescue. An MSI presentation at the 2022 American Evaluation Association annual meeting.
Analysis of Variance
Tutorial on the ANOVA theorem
Expectation Maximization
Exploration and tutorial on the Expectation Maximization (EM) algorithm
Palestinian Perception Survey - Evaluation Question 1
Weighted frequency output of the Palestian Perception Survey
Introduction to Causal Inference
A gentle introduction to the field
South Sudan resilience
Frequencies of the first South Sudan resilience panel survey
USAID evaluation trends
Count and quality of USAID evaluations over time
A tutorial on making a map in R
A brief tutorial on making a map in both tmap and ggplot2. This approach uses sf for working with vector data to make a static map and an interactive map.
Using design effects to help plan survey sample sizes
Illustration of how we might conduct complex survey sample size estimations using information from similar surveys