Course Introduction
The Regression Discontinuity and Advanced Programme Evaluation Training Course provides a rigorous and practical framework for evaluating programmes, policies, interventions, and public initiatives using regression discontinuity and complementary quasi-experimental methods. Participants will develop advanced knowledge of causal inference, treatment assignment mechanisms, counterfactual reasoning, programme effects, and the conditions under which discontinuities in eligibility or treatment can generate credible causal estimates.
The course provides detailed coverage of sharp and fuzzy regression discontinuity designs, with particular attention to cutoff-based treatment assignment, local treatment effects, continuity assumptions, bandwidth selection, functional-form choices, and statistical inference. Participants will learn how to identify appropriate running variables and thresholds, assess whether observations around a cutoff are comparable, and determine whether the institutional assignment mechanism provides a defensible basis for causal identification.
Advanced programme evaluation concepts are integrated throughout the course so that participants can connect statistical methods with practical evaluation questions. The programme addresses evaluation design, theory of change, treatment exposure, outcome measurement, implementation fidelity, counterfactual construction, impact estimands, heterogeneous effects, and interpretation of evidence. Participants will learn to select appropriate evaluation strategies according to programme characteristics, available data, implementation structures, and research objectives.
Significant attention is given to threats to the validity of regression discontinuity studies, including manipulation around the cutoff, sorting, imperfect compliance, discontinuities in covariates, bandwidth sensitivity, functional-form misspecification, measurement error, interference, treatment anticipation, and concurrent changes at the threshold. Participants will apply graphical diagnostics, density tests, covariate continuity assessments, placebo cutoffs, alternative specifications, and sensitivity analyses to evaluate the credibility of empirical findings.
The course also explores advanced programme-evaluation approaches that complement regression discontinuity, including difference-in-differences, matching, instrumental variables, interrupted time series, synthetic controls, propensity-score methods, and longitudinal evaluation. Participants will learn how these approaches differ in their assumptions and interpretation, how they can be combined where appropriate, and how to develop coherent evaluation strategies rather than selecting statistical techniques solely on the basis of data availability.
By the end of the course, participants will be able to design, implement, diagnose, interpret, and communicate sophisticated regression discontinuity and programme-evaluation studies. Emerging topics including heterogeneous treatment effects, machine-learning-assisted evaluation, administrative and high-frequency data, geographic discontinuity designs, regression kink designs, reproducible workflows, transparent reporting, and modern causal-inference methods will prepare participants to address increasingly complex programme-evaluation challenges.
Duration
10 days
Who Should Attend
-
Programme evaluation specialists responsible for assessing intervention effectiveness and implementation outcomes.
-
Monitoring and evaluation professionals seeking advanced causal inference and impact-evaluation capabilities.
-
Economists and econometricians conducting empirical programme and policy research.
-
Government analysts evaluating public programmes, eligibility systems, subsidies, grants, and social interventions.
-
Development practitioners responsible for designing or reviewing rigorous programme evaluations.
-
Policy analysts assessing the effects of regulations, public services, eligibility rules, and government interventions.
-
Researchers working with administrative, survey, longitudinal, programme, or institutional datasets.
-
Data scientists and statisticians interested in advanced quasi-experimental evaluation methods.
-
Consultants providing programme evaluation, economic analysis, impact assessment, and evidence-generation services.
-
Academic researchers developing applied causal-inference studies and evaluation research.
-
Programme managers seeking to interpret evaluation evidence and understand causal programme impacts.
-
Professionals preparing technical evaluation reports, research publications, policy briefs, and evidence-based programme recommendations.
Course Objectives
-
Develop advanced knowledge of regression discontinuity methodology and its application to programme, policy, and intervention evaluation.
-
Understand the counterfactual framework, causal estimands, and identification assumptions underlying regression discontinuity designs.
-
Distinguish between sharp and fuzzy regression discontinuity designs and determine when each approach is appropriate for evaluation.
-
Assess continuity and comparability around assignment thresholds using graphical, statistical, institutional, and contextual evidence.
-
Apply appropriate bandwidth-selection, local-regression, polynomial, functional-form, and inference strategies for regression discontinuity analysis.
-
Detect manipulation, sorting, bunching, and other forms of strategic behavior around programme eligibility thresholds and treatment cutoffs.
-
Conduct advanced validity checks using covariate continuity, density testing, placebo thresholds, alternative bandwidths, and specification sensitivity.
-
Interpret local treatment effects appropriately and assess the extent to which estimated effects can inform broader programme or population-level conclusions.
-
Integrate regression discontinuity with complementary programme-evaluation approaches including DiD, IV, matching, interrupted time series, and synthetic controls.
-
Address practical evaluation challenges involving imperfect compliance, treatment contamination, spillovers, missing data, measurement error, and concurrent interventions.
-
Develop reproducible evaluation workflows covering data preparation, estimation, diagnostics, robustness analysis, visualization, interpretation, and transparent reporting.
-
Design complete programme-evaluation studies that connect evaluation questions, theory of change, identification strategy, empirical analysis, findings, and evidence-based conclusions.
Comprehensive Course Outline
Module 1: Foundations of Advanced Programme Evaluation
-
Principles of programme evaluation, causal inference, counterfactual reasoning, treatment effects, and evidence-based assessment of interventions.
-
Theory of change, logic models, programme objectives, implementation pathways, outcome measurement, and evaluation questions.
-
Experimental, quasi-experimental, observational, process, outcome, and impact evaluation designs and their different analytical purposes.
-
Defining causal estimands, target populations, treatment exposure, comparison groups, evaluation periods, and measurable programme outcomes.
Module 2: Regression Discontinuity Design Fundamentals
-
Foundations of regression discontinuity based on treatment assignment determined by a known threshold on an observed running variable.
-
Sharp regression discontinuity designs where treatment status changes deterministically at the programme eligibility or assignment cutoff.
-
Fuzzy regression discontinuity designs where crossing the threshold changes treatment probability rather than guaranteeing treatment receipt.
-
Local identification, counterfactual reasoning, and interpretation of treatment effects for observations located near the assignment threshold.
Module 3: Assignment Mechanisms, Running Variables, and Cutoffs
-
Identifying running variables, eligibility scores, administrative thresholds, geographic boundaries, dates, and other assignment mechanisms suitable for discontinuity designs.
-
Understanding institutional rules governing treatment assignment and translating administrative procedures into formal empirical identification strategies.
-
Assessing cutoff validity, threshold precision, eligibility definitions, measurement units, and potential changes in assignment rules over time.
-
Identifying multiple thresholds, nested eligibility rules, overlapping programmes, and complex assignment systems that can affect causal interpretation.
Module 4: Identification Assumptions and Local Comparability
-
Continuity of potential outcomes around the cutoff and the role of local comparability in establishing a credible counterfactual.
-
Assessing whether unobserved determinants of outcomes can reasonably be expected to vary smoothly through the assignment threshold.
-
Examining covariate continuity, baseline characteristics, demographic composition, institutional characteristics, and pre-treatment outcomes around the cutoff.
-
Identifying threats caused by simultaneous policy changes, discontinuous institutional conditions, sorting, manipulation, and measurement problems.
Module 5: Estimation and Bandwidth Selection
-
Local linear and local polynomial regression approaches for estimating discontinuities and treatment effects around programme thresholds.
-
Bandwidth selection, bias-variance trade-offs, data density, effective sample size, and implications for local treatment-effect estimation.
-
Comparing conventional, robust bias-corrected, and alternative inference procedures for regression discontinuity applications.
-
Assessing sensitivity to bandwidth choices, polynomial orders, kernels, covariate adjustment, and alternative local specifications.
Module 6: Manipulation, Sorting, and Density Testing
-
Understanding strategic manipulation, eligibility gaming, administrative sorting, and behavioral responses around programme assignment thresholds.
-
Graphical assessment of the distribution of observations around the cutoff and identification of unusual bunching or discontinuities in density.
-
Formal density-based testing and complementary institutional investigations for detecting manipulation of the running variable.
-
Addressing measurement precision, heaping, rounding, administrative discretion, and other mechanisms that can produce apparent discontinuities.
Module 7: Validity Checks and Diagnostic Testing
-
Testing continuity of predetermined covariates, baseline characteristics, and other variables that should not change discontinuously at the treatment threshold.
-
Placebo cutoff tests using artificial thresholds to determine whether similar discontinuities appear where no treatment change occurs.
-
Alternative outcome measures, pre-treatment outcomes, negative-control variables, and falsification tests for strengthening design credibility.
-
Graphical diagnostics, residual analysis, influential observations, bandwidth sensitivity, and systematic documentation of specification choices.
Module 8: Advanced Regression Discontinuity Designs
-
Regression discontinuity designs with multiple treatment levels, multiple cutoffs, repeated thresholds, and complex eligibility rules.
-
Geographic regression discontinuity using administrative boundaries, geographic borders, service areas, and spatial assignment mechanisms.
-
Regression kink designs where treatment intensity changes in slope rather than through a discrete level shift at a threshold.
-
Designs involving discrete running variables, coarsened scores, limited support, measurement error, and imperfect threshold precision.
Module 9: Imperfect Compliance and Treatment Effects
-
Understanding imperfect treatment take-up, noncompliance, treatment refusal, programme leakage, and deviations from eligibility-based assignment.
-
Instrumental-variable interpretation of fuzzy regression discontinuity and the relationship between eligibility, treatment receipt, and local causal effects.
-
Estimating local average treatment effects under imperfect compliance while carefully considering relevance and exclusion assumptions.
-
Evaluating heterogeneous compliance, treatment switching, partial exposure, and differential programme participation around the cutoff.
Module 10: Advanced Programme Evaluation Methodologies
-
Difference-in-differences for evaluating policy changes over time and comparing treated and untreated groups under appropriate trend assumptions.
-
Instrumental variables, matching, propensity-score methods, interrupted time series, and synthetic controls as complementary evaluation strategies.
-
Selecting evaluation designs based on assignment mechanisms, available data, treatment timing, programme structure, and the intended causal estimand.
-
Combining quasi-experimental approaches where justified while maintaining clear identification arguments and avoiding unsupported causal claims.
Module 11: Heterogeneous Treatment Effects and Programme Equity
-
Examining treatment-effect heterogeneity across demographic groups, geographic areas, baseline characteristics, risk levels, and programme exposure.
-
Evaluating differential programme impacts by eligibility status, treatment intensity, implementation quality, institutional setting, and participant characteristics.
-
Assessing equity implications through subgroup analysis while addressing multiple testing, statistical power, selective reporting, and interpretation challenges.
-
Exploring causal machine-learning approaches for discovering heterogeneous effects while maintaining transparent research design and appropriate validation.
Module 12: Data Quality, Missingness, and Measurement Challenges
-
Handling missing observations, incomplete programme records, participant attrition, unbalanced datasets, and irregular administrative reporting.
-
Measurement error in running variables, treatment status, programme participation, outcomes, eligibility scores, and administrative classifications.
-
Data linkage, duplicate records, changing variable definitions, coding inconsistencies, geographic identifiers, and longitudinal programme-record challenges.
-
Applying imputation, weighting, sensitivity analysis, data-quality checks, and transparent documentation of assumptions concerning incomplete information.
Module 13: Statistical Inference, Robustness, and Sensitivity Analysis
-
Selecting appropriate standard-error and inference procedures for local regression discontinuity designs with clustering, heteroskedasticity, and limited effective samples.
-
Robustness analysis using alternative bandwidths, polynomial specifications, kernels, covariates, outcome definitions, and estimation procedures.
-
Sensitivity analysis for unobserved confounding, manipulation, measurement error, functional-form assumptions, and departures from continuity.
-
Distinguishing statistical uncertainty from uncertainty created by identification assumptions, programme implementation, design limitations, and external validity.
Module 14: Programme Implementation and Impact Interpretation
-
Connecting estimated programme effects with implementation fidelity, treatment intensity, participant behavior, institutional capacity, and programme delivery conditions.
-
Interpreting statistical significance, effect sizes, confidence intervals, practical significance, uncertainty, and local treatment effects for programme stakeholders.
-
Distinguishing short-term outcomes, intermediate results, long-term impacts, unintended consequences, and implementation-related programme effects.
-
Assessing external validity, generalizability, scalability, transportability, and the limitations of applying local threshold effects to broader populations.
Module 15: Modern Evaluation Analytics and Emerging Methods
-
Applying machine-learning techniques to flexible covariate adjustment, heterogeneous treatment effects, prediction, and high-dimensional programme-evaluation datasets.
-
Using administrative big data, real-time programme monitoring data, geospatial information, digital records, and high-frequency observations for evaluation.
-
Emerging approaches involving automated causal workflows, advanced discontinuity estimators, generalized designs, privacy-preserving data integration, and reproducible analytical systems.
-
Addressing transparency, research credibility, open analytical workflows, preregistration, replication, computational reproducibility, and responsible use of advanced evaluation technologies.
Module 16: Integrated Regression Discontinuity and Programme Evaluation Projects
-
Designing a complete regression discontinuity evaluation from programme theory and research question through identification, data preparation, estimation, diagnostics, and reporting.
-
Developing integrated evaluation strategies that combine regression discontinuity with complementary quasi-experimental approaches when research conditions justify their use.
-
Producing reproducible analytical workflows with documented data processing, model specifications, robustness checks, visualizations, and clearly stated assumptions.
-
Preparing professional evaluation reports that communicate methods, causal evidence, limitations, implementation context, uncertainty, and policy-relevant findings transparently.
Training Approach
This course will be delivered by our skilled trainers who have vast knowledge and experience as expert professionals in the fields. The course is taught in English and through a mix of theory, practical activities, group discussion and case studies. Course manuals and additional training materials will be provided to the participants upon completion of the training.
Tailor-Made Course
This course can also be tailor-made to meet organization requirement. For further inquiries, please contact us on: Email: training@upskilldevelopment.com Tel: +254 721 331 808
Training Venue
The training will be held at our Upskill Training Centre. We also offer training for a group (at a discount of 10% to 50%) at requested location all over the world. The Onsite course fee covers the course tuition, training materials, two break refreshments, buffet lunch, airport transfers, Upskill gift package, and guided tour.
Visa application, travel expenses, dinners, accommodation, insurance, and other personal expenses are catered by the participant
Certification
Participants will be issued with Upskill certificate upon completion of this course.
Airport Pickup and Accommodation
Airport pickup and accommodation is arranged upon request. For booking contact our Training Coordinator through Email: training@upskilldevelopment.com, +254 721 331 808
Terms of Payment:
Unless otherwise agreed between the two parties’ payment of the course fee should be done 3 working days before commencement of the training so as to enable us to prepare better.
