Course Introduction
Longitudinal Data Management and Repeated Measures Analysis is an advanced professional training course designed to strengthen participants’ ability to manage, analyse and interpret data collected repeatedly from the same individuals, households, organisations or communities over time. It combines practical data management techniques with statistical methods for examining change, trajectories, intervention effects and evolving relationships. Participants learn to develop reliable longitudinal datasets, preserve participant histories across measurement rounds and select appropriate analytical methods for monitoring development programmes, evaluating interventions and conducting applied social research.
Longitudinal studies are particularly valuable when programme outcomes develop gradually or when decision-makers need to understand whether observed improvements are sustained. A baseline assessment may establish participants’ initial conditions, while subsequent follow-ups document changes in employment, household income, educational attainment, health, financial inclusion or access to essential services. However, differences in interview timing, inconsistent identifiers, changes in questionnaire design and incomplete follow-up can undermine the comparability of observations. The course shows participants how to establish robust data structures, harmonise repeated measurements and document the processes required to maintain analytical consistency across survey rounds.
A major component of the programme focuses on longitudinal data management, including participant identification, record linkage, reshaping datasets between wide and long formats, date handling, variable harmonisation and data quality assurance. Participants learn to distinguish genuine changes in participant status from data entry errors, duplicate records, revised classifications and inconsistent measurement units. They also examine how to preserve a complete history of enrolment, exposure, participation, outcome measurement and follow-up. Through applied exercises, participants develop data dictionaries, tracking systems and reproducible data preparation workflows that support credible analysis and efficient collaboration between research, monitoring and programme management teams.
The statistical component introduces repeated measures analysis, longitudinal regression, linear and generalised linear mixed-effects models, generalised estimating equations and growth curve modelling. Participants investigate within-person change, between-person differences, time trends, nonlinear trajectories and differences in outcomes between intervention and comparison groups. They learn to select appropriate covariance structures, interpret random effects and fixed effects, and test time-by-group interactions that reveal whether outcome trajectories differ between programme participants and comparison populations. The course also explains why analytical choices must reflect the research question, measurement schedule, outcome distribution and structure of the available data.
Missing observations and participant attrition are addressed throughout the training because they can materially affect longitudinal conclusions. Participants learn to describe follow-up patterns, distinguish intermittent missing values from permanent dropout, investigate plausible missing-data mechanisms and apply defensible approaches such as likelihood-based estimation and multiple imputation where appropriate. They also examine sensitivity analysis, inverse probability weighting and the limitations of complete-case analysis. Emphasis is placed on recognising that sophisticated statistical methods cannot automatically eliminate bias when participants with missing observations differ systematically from those who remain in a study.
The final emphasis is on converting longitudinal evidence into practical programme learning and accountable decision-making. Participants learn to produce trajectory plots, descriptive summaries, model-based predictions and clear analytical reports that communicate changes over time alongside appropriate measures of uncertainty. Applied case studies draw on employment programmes, agricultural livelihoods, education, public health, social protection and financial inclusion. By the end of the course, participants will be able to establish reproducible longitudinal data workflows, apply appropriate repeated measures methods and communicate the strength and limitations of evidence about programme performance, sustained outcomes and changes in beneficiary wellbeing.
Duration
10 days
Who Should Attend
-
Monitoring, Evaluation and Learning specialists responsible for baseline, midline, endline and follow-up assessments.
-
Development evaluation professionals analysing programme outcomes across multiple reporting periods.
-
Research officers managing panel surveys, longitudinal cohorts and repeated beneficiary assessments.
-
Data managers responsible for participant tracking, record linkage, data harmonisation and longitudinal database maintenance.
-
Quantitative analysts working with repeated observations from individuals, households, institutions or communities.
-
Programme managers seeking to understand whether beneficiary outcomes improve, deteriorate or remain stable over time.
-
Economists and social scientists investigating income trajectories, employment transitions and changes in household welfare.
-
Public health researchers analysing repeated clinical, behavioural, service utilisation or community health measurements.
-
Education researchers examining student progress, school participation and learning outcomes across academic periods.
-
Agricultural development specialists evaluating changes in productivity, technology adoption and household livelihoods over successive seasons.
-
Government statisticians and policy analysts working with panel datasets and repeated administrative records.
-
Donor agency advisers reviewing longitudinal evaluation designs, attrition management and evidence of sustained impact.
-
NGO research and evaluation teams conducting multi-round surveys and tracking beneficiary outcomes after programme completion.
-
Academic researchers, postgraduate students and consultants seeking practical competence in longitudinal statistical methods.
Course Objectives
-
Explain the principles of longitudinal research and repeated measures analysis, including the advantages and limitations of following the same units over multiple time points.
-
Design reliable longitudinal data structures that preserve participant identities, observation dates, measurement occasions and the relationships between records.
-
Apply data cleaning, record linkage, variable harmonisation and consistency checks to maintain data quality across repeated survey rounds and administrative extracts.
-
Distinguish within-person change from between-person differences and select statistical approaches that appropriately represent both sources of variation.
-
Fit and interpret repeated measures models, linear mixed-effects models and generalised estimating equations for suitable outcome types and research questions.
-
Specify and interpret fixed effects, random intercepts, random slopes, time effects and group-by-time interactions in longitudinal programme evaluation.
-
Select suitable covariance structures and assess model fit, residual behaviour and assumptions affecting repeated measures inference.
-
Examine missing-data patterns, participant attrition and follow-up differences, applying appropriate methods and sensitivity analyses where justified.
-
Estimate and communicate outcome trajectories, predicted changes, subgroup differences and uncertainty measures using tables, graphs and model-based summaries.
-
Integrate longitudinal methods into programme impact evaluation while distinguishing observed change, adjusted associations and credible causal effects.
-
Implement reproducible longitudinal data management and statistical analysis workflows using R and other appropriate statistical software.
-
Translate longitudinal results into actionable recommendations for programme adaptation, sustained outcome monitoring, reporting and evidence-informed resource allocation.
Comprehensive Course Outline
Module 1: Foundations of Longitudinal Research and Repeated Measures
-
Explain longitudinal research designs and distinguish panel studies, cohort studies, repeated cross-sectional surveys and other approaches to measuring change over time.
-
Identify the analytical benefits of repeated observations for studying trajectories, temporal ordering, sustained outcomes and differences in programme response.
-
Distinguish within-person variability, between-person variability, measurement error and changes in population composition across successive observation periods.
-
Formulate longitudinal evaluation questions, specify estimands and link research objectives to measurement schedules, data requirements and appropriate statistical methods.
Module 2: Longitudinal Study Design and Measurement Planning
-
Develop measurement schedules that align baseline, implementation, follow-up and post-programme assessments with the expected timing of programme outcomes.
-
Define consistent outcome measures, explanatory variables, exposure indicators and participant characteristics that can be compared across study waves.
-
Assess the implications of irregular observation intervals, changes in measurement instruments and different follow-up periods for analytical design.
-
Prepare longitudinal data protocols covering participant tracking, consent, fieldwork procedures, questionnaire versions, metadata and quality assurance responsibilities.
Module 3: Participant Identification, Record Linkage and Data Architecture
-
Develop stable participant identifiers and linkage rules that support accurate matching across survey rounds, administrative systems and programme databases.
-
Detect duplicate records, inconsistent identifiers, split or merged household histories and other linkage problems that may distort longitudinal analysis.
-
Establish relational data structures linking participants, household members, visits, interventions, outcomes and follow-up events without losing observation history.
-
Document linkage quality, unresolved matches, identifier changes and data provenance to ensure that longitudinal datasets remain auditable and reproducible.
Module 4: Longitudinal Data Cleaning and Harmonisation
-
Reshape datasets between wide and long formats and determine which representation is appropriate for particular management tasks and statistical models.
-
Harmonise variable names, coding schemes, response categories, measurement units and derived indicators across repeated rounds of data collection.
-
Identify impossible dates, inconsistent chronology, unexpected changes, duplicate observations and implausible values using systematic validation rules.
-
Maintain data dictionaries, transformation logs, version control and documented correction procedures so that each analytical dataset can be reconstructed.
Module 5: Exploratory Analysis and Visualisation of Change
-
Summarise participant characteristics, outcome distributions and observation counts at each measurement occasion before fitting longitudinal models.
-
Plot individual trajectories, group-level means and distributions over time to identify trends, heterogeneity, outliers and potential data quality concerns.
-
Examine follow-up participation patterns and compare baseline characteristics between participants with complete and incomplete observation histories.
-
Develop an initial longitudinal data analysis report that documents missingness, variability, timing differences and implications for the planned statistical analysis.
Module 6: Repeated Measures ANOVA and General Linear Approaches
-
Explain the assumptions of repeated measures ANOVA and identify circumstances in which it may be suitable for balanced, regularly scheduled observations.
-
Test time effects and group-by-time interactions to investigate whether outcomes change and whether change patterns differ across programme groups.
-
Assess sphericity, covariance assumptions, missing observations and unequal follow-up, recognising the limitations of traditional repeated measures procedures.
-
Compare repeated measures ANOVA with mixed-effects and other longitudinal methods to select approaches appropriate to the data structure and evaluation objective.
Module 7: Linear Mixed-Effects Models for Longitudinal Outcomes
-
Specify linear mixed-effects models for continuous outcomes and distinguish population-level fixed effects from participant-specific random effects.
-
Interpret random intercepts and random slopes to investigate differences in baseline outcome levels and rates of change across participants or clusters.
-
Model time using categorical indicators, continuous trends or suitable nonlinear terms according to the measurement schedule and expected trajectory.
-
Compare candidate model specifications and covariance structures using diagnostics, information criteria and substantive understanding of the longitudinal process.
Module 8: Generalised Estimating Equations and Correlated Outcomes
-
Explain the population-averaged interpretation of generalised estimating equations and distinguish it from subject-specific mixed-effects modelling.
-
Specify appropriate link functions and outcome distributions for repeated binary, count and other non-continuous response variables.
-
Select working correlation structures and interpret robust standard errors while considering sample size, cluster counts and missing-data assumptions.
-
Compare GEE and mixed-effects approaches in terms of estimands, interpretation, assumptions and suitability for the intended programme evaluation question.
Module 9: Growth Curve Modelling and Nonlinear Trajectories
-
Develop growth curve models to estimate initial outcome levels, average rates of change and individual differences in longitudinal trajectories.
-
Compare linear, quadratic, piecewise and other substantively justified time functions when programme outcomes may change at different rates.
-
Examine differences in trajectory shape across beneficiary groups, implementation settings or exposure categories using suitable interaction terms.
-
Interpret predicted trajectories, uncertainty intervals and individual variation to identify when outcomes improve, plateau, decline or show delayed responses.
Module 10: Programme Evaluation and Treatment Effects Over Time
-
Specify longitudinal evaluation models that compare changes in intervention and comparison groups across baseline and follow-up measurement occasions.
-
Interpret group-by-time interactions and difference-in-differences estimates while distinguishing descriptive change from effects supported by a credible research design.
-
Assess baseline comparability, differential attrition, changes in programme exposure and other threats to the validity of longitudinal impact estimates.
-
Conduct robustness checks and sensitivity analyses to investigate whether conclusions depend on outcome definitions, time specifications or analytical assumptions.
Module 11: Missing Data, Attrition and Incomplete Follow-Up
-
Distinguish intermittent missing measurements, item nonresponse, permanent dropout and administrative loss to follow-up within longitudinal datasets.
-
Explain missing completely at random, missing at random and missing not at random assumptions and their implications for estimation and interpretation.
-
Apply appropriate likelihood-based methods or multiple imputation when their assumptions and implementation are suitable for the study design and missingness patterns.
-
Evaluate complete-case analysis, inverse probability weighting and sensitivity analysis while documenting the potential influence of non-random attrition.
Module 12: Advanced Covariance Structures and Model Diagnostics
-
Examine compound symmetry, autoregressive, unstructured and other covariance structures to represent dependence between repeated observations.
-
Assess residual distributions, heteroscedasticity, influential participants, convergence problems and other diagnostics relevant to longitudinal model validity.
-
Compare model specifications using appropriate likelihood-based tests and information criteria, recognising the limitations of mechanical model selection.
-
Investigate the sensitivity of estimates and standard errors to alternative time representations, random-effects structures and residual correlation assumptions.
Module 13: Time-Varying Exposures, Confounding and Advanced Methods
-
Distinguish baseline covariates from time-varying exposures and identify how changes in participation, service intensity or contextual conditions relate to outcomes.
-
Investigate time-varying confounding and explain why conventional regression adjustment may be inadequate when previous exposure affects later confounders.
-
Introduce marginal structural models and inverse probability weighting as advanced approaches for selected longitudinal causal questions when assumptions and data support them.
-
Assess temporal ordering, reverse causation and measurement timing, and explain why repeated observations alone do not establish causal relationships.
Module 14: Practical Longitudinal Data Analysis Using R
-
Import longitudinal datasets, validate participant identifiers and observation dates, and create analysis-ready data structures with reproducible preparation scripts.
-
Use appropriate R tools, including lme4, nlme and geepack where relevant, to estimate mixed-effects models and generalised estimating equations.
-
Generate descriptive trajectory plots, model predictions, confidence intervals and diagnostic summaries to support interpretation of longitudinal findings.
-
Organise scripts, package versions, data dictionaries and model outputs into a transparent analytical workflow that facilitates independent review and replication.
Module 15: Longitudinal Reporting, Visualisation and Programme Learning
-
Present changes over time using clear tables, trajectory plots and model-based estimates that distinguish observed averages from adjusted predictions.
-
Communicate effect sizes, uncertainty intervals, sample sizes and attrition patterns so that decision-makers can judge the strength of longitudinal evidence.
-
Translate findings into programme recommendations while distinguishing short-term responses, sustained improvements, delayed effects and potentially adverse trajectories.
-
Document limitations arising from measurement changes, missing data, selection, model assumptions and incomplete follow-up in technical and policy-oriented reports.
Module 16: Integrated Capstone and Longitudinal Quality Assurance
-
Develop a complete longitudinal analysis plan connecting research questions, participant tracking, outcome definitions, missing-data strategies and statistical methods.
-
Complete an applied case study involving data linkage, harmonisation, exploratory analysis, repeated measures modelling and interpretation of change over time.
-
Prepare a reproducible technical report with model specifications, diagnostic evidence, trajectory visualisations, uncertainty estimates and transparent limitations.
-
Present findings to a review panel and defend analytical decisions while identifying priorities for improving data collection, retention and future programme evaluation.
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.








