R Statistical Analysis for Programme Evaluation Training Course

Course overview

Programme teams are increasingly expected to demonstrate not only that activities were delivered, but whether interventions contributed to measurable changes for intended populations. Yet monitoring data often sits in spreadsheets, baseline and endline datasets are inconsistently structured, and evaluation findings can be difficult to reproduce or defend. This course equips development and humanitarian professionals to use R to clean, analyse and communicate evidence from programme monitoring systems, surveys, assessments and evaluations.

Participants learn an end-to-end analytical workflow in R and RStudio: importing and validating data, reshaping records, managing missing values, creating indicators, producing disaggregated descriptive statistics, testing group differences, estimating regression models and interpreting results in relation to a theory of change. The course uses tidyverse workflows for data preparation, survey-aware analysis principles, reproducible scripts, publication-ready visualisations and Quarto reporting. Particular attention is given to common M&E questions, including change over time, outcome comparisons, targeting analysis, equity disaggregation and data-quality checks.

Instruction combines short technical demonstrations with guided coding labs based on realistic programme datasets such as household surveys, post-distribution monitoring and baseline/endline studies. Participants build an annotated R project containing reusable scripts, a cleaned analysis dataset, indicator tables, charts and a Quarto evaluation findings report. This provides a practical template they can adapt for their own programme portfolio after the course.

The course is designed for M&E staff, evaluation specialists, programme analysts and research personnel who already work with quantitative programme data and need a robust, transparent alternative to spreadsheet-only analysis. Managers gain staff who can produce more consistent evidence products, document analytical decisions and interrogate results before they inform programme adaptation or donor reporting.

Course objectives

By the end of this course, participants will be able to:

• Build a structured RStudio project for a programme evaluation, including scripts, source data, outputs and documented assumptions
• Import, inspect and validate monitoring, survey and baseline/endline datasets using R data-quality checks
• Clean, reshape and join programme datasets with tidyverse functions such as dplyr, tidyr and stringr
• Construct outcome indicators, disaggregation variables and analysis-ready denominator definitions from questionnaire data
• Produce weighted and unweighted descriptive statistics, confidence intervals and cross-tabulations for M&E reporting
• Test differences between participant groups or survey rounds using appropriate statistical tests and interpret the results
• Estimate and explain linear and logistic regression models for programme outcome analysis
• Generate a reproducible Quarto evaluation report containing tables, visualisations, methods notes and findings

Benefits of attending

Target competencies

Who should attend

• Monitoring and Evaluation Officers — who need to turn routine and survey data into defensible programme findings
• Evaluation Specialists — who analyse baseline, endline and follow-up datasets and must document reproducible methods
• Programme Managers — who need to interrogate outcome evidence before adapting activities or reporting to donors
• Research and Learning Officers — who design quantitative studies and require transparent analytical workflows
• Data Analysts in NGOs and UN agencies — who support indicator reporting, assessments and evidence products
• Humanitarian Information Management Officers — who prepare and analyse needs assessment and response-monitoring data

Requirements and prerequisites

Participants should be comfortable working with tabular data in Excel or CSV format and understand core M&E concepts such as indicators, numerators and denominators, disaggregation, baseline/endline measurement, sampling and basic descriptive statistics. Prior exposure to R is helpful but not essential; the course begins with RStudio navigation and guided coding patterns before progressing to statistical models. Participants should bring a laptop on which they can install R and RStudio, subject to organisational IT policy. Advanced mathematics, prior programming experience and prior use of regression software are not required, although participants should be prepared to work directly with code.

Training methodology

Each day combines instructor-led explanation of an M&E analytical problem with live RStudio demonstrations and closely guided coding labs. Participants work with realistic development and humanitarian datasets, including household survey records, beneficiary registers and baseline/endline extracts. Exercises require participants to diagnose data issues, write and run R scripts, interpret outputs and discuss the programme implications of findings. Small-group reviews focus on choices such as denominators, disaggregation and model interpretation. The final day includes an application-planning workshop in which participants adapt a reporting workflow to a live or anticipated evaluation task.

Course outline

Day 1: R foundations for evaluation data

• RStudio projects, scripts, consoles and working-directory discipline
• Importing CSV and Excel monitoring datasets with readr and readxl
• Inspecting variable types, labels, ranges and duplicate identifiers
• Tidy data principles for survey and programme records
• R object types, vectors, data frames and factors
• Documenting analytical decisions with comments and code structure
• Initial data-quality checks for completeness, validity and uniqueness

Workshop: Participants create an RStudio evaluation project, import a household survey dataset and produce a documented data-quality log identifying priority issues.

Day 2: Data preparation and indicator construction

• Filtering, selecting, mutating and summarising data with dplyr
• Recoding categorical responses and managing labelled variables
• Handling missing, refused and not-applicable survey responses
• Reshaping wide and long datasets with pivot_longer and pivot_wider
• Joining beneficiary, activity and survey datasets using unique keys
• Constructing numerators, denominators and binary outcome indicators
• Creating sex, age, disability, location and vulnerability disaggregations

Workshop: Participants clean and merge programme registration and endline survey files, then generate a documented indicator dataset with agreed denominators.

Day 3: Descriptive analysis and visual evidence

• Frequency tables and cross-tabulations for monitoring indicators
• Means, medians, proportions and distribution diagnostics
• Confidence intervals for proportions and mean estimates
• Weighted versus unweighted estimates and survey-design considerations
• Comparing baseline, endline and monitoring rounds
• Building bar charts, trend charts and distribution plots with ggplot2
• Applying clear disaggregation and uncertainty conventions in reporting

Workshop: Participants produce a disaggregated indicator table and a set of ggplot2 charts for an endline briefing, with written interpretation of key patterns.

Day 4: Statistical testing and explanatory models

• Formulating evaluation questions, hypotheses and comparison groups
• Chi-square and Fisher's exact tests for categorical outcomes
• T-tests and non-parametric tests for continuous and ordinal measures
• Effect sizes, p-values and confidence intervals in evaluation interpretation
• Linear regression for continuous programme outcome measures
• Logistic regression for binary outcome indicators
• Checking model assumptions, confounding and limitations of observational data

Workshop: Participants test outcome differences across participant groups and fit a logistic regression model, then write a plain-language findings note for a programme manager.

Day 5: Reproducible reporting and evaluation application

• Quarto document structure, code chunks and parameterised reporting
• Automating tables and figures from cleaned analysis datasets
• Writing methods notes, limitations and interpretation statements
• Creating publication-ready outputs in HTML, Word and PDF formats
• Version control principles and file-management standards for evaluation teams
• Peer review of R scripts, indicator definitions and analytical outputs
• Translating statistical findings into programme recommendations and learning questions

Workshop: Participants complete a Quarto evaluation report containing automated tables, charts, methods notes and recommendations, then develop an action plan for applying the workflow to their own work.

Tools & standards covered

R, RStudio, tidyverse, Quarto

A typical training day

Live online deliveries follow the same structure in the East Africa Time zone, with shorter screen blocks and longer breaks.

What the fee includes

How you can take this course

Certification

Participants who complete the full five days receive the Skillset Development Certificate of Completion, stating the course title, course code, dates and delivery format — suitable for professional-development records and employer reimbursement.

Frequently asked questions