Digital Technologies for Monitoring and Evaluation Training Course

Participants can specify secure, reliable mobile data collection workflows for monitoring and evaluation.

• Form logic uses skip patterns, constraints, calculations and required fields
• Offline mobile workflows manage synchronization, enumerator accounts and device allocation
• KoboToolbox forms configured with multilingual labels, consent and validation rules
• Computer-assisted personal interviewing reduces transcription errors and missing values
• Paradata records interview duration, GPS points, edits and synchronization history
• Digital consent scripts document voluntary participation, permissions and withdrawal routes

Participants can define data structures and interoperability requirements for connected M&E systems.

• Unique identifiers link participant records, services, outputs and outcome observations
• Data models distinguish transactional, longitudinal, geospatial and unstructured information
• DHIS2 metadata structures organise indicators, reporting periods and organisational units
• Application programming interfaces support documented transfers between approved systems
• Interoperability mapping exposes duplicate fields, incompatible definitions and missing ownership
• Data dictionaries assign formats, permissible values, custodians and retention periods

Participants can assess data quality and establish proportionate controls for sensitive M&E data.

• Routine data quality assessment measures completeness, timeliness, consistency and accuracy
• Validation rules flag impossible values, duplicates, outliers and denominator errors
• Audit trails preserve corrections, approvals, timestamps and accountable user identities
• Role-based access controls restrict sensitive records according to documented responsibilities
• Data protection impact assessments identify harms, safeguards and residual risk
• Encryption, backups and incident response protect monitoring systems from disruption
• Bias audits test coverage gaps affecting women, displaced people and minorities

Participants can use spatial data and mapping methods to analyse coverage, access and equity.

• Coordinate reference systems prevent positional errors when combining spatial datasets
• QGIS joins link survey records, administrative boundaries and service locations
• Choropleth maps communicate rates using suitable denominators and classification methods
• Spatial sampling targets underserved areas while documenting access and security constraints
• Hotspot analysis distinguishes clustering from population density and reporting artefacts
• Participatory mapping validates facility locations, hazards and community-defined boundaries

Participants can judge when remotely collected data provides valid evidence for M&E questions.

• Satellite imagery provides proxy measures for land cover, damage and access
• Normalized Difference Vegetation Index tracks vegetation change during livelihood monitoring
• Change detection compares dated imagery using documented thresholds and ground truthing
• Remote sensing validation combines reference samples, confusion matrices and accuracy metrics
• IoT sensor data requires calibration, maintenance logs and plausible operating ranges
• Drone imagery protocols address permissions, safety, privacy and georeferenced metadata

Participants can evaluate AI-supported analysis for suitability, reliability, bias and human oversight.

• Machine learning problem formulation separates prediction, classification and causal inference
• Training and test splits assess generalisation without contaminating evaluation evidence
• Natural language processing codes feedback themes from open-text responses
• Human review samples detect hallucinations, misclassification and harmful automated decisions
• Algorithmic bias assessment examines labels, representation, error rates and disparate impacts
• Model cards document intended use, limitations, performance and oversight requirements
• Generative AI prompts protect confidential data and preserve analyst accountability

Participants can specify decision-focused dashboards and visualisations that communicate evidence accurately.

• Dashboard requirements begin with decisions, users, refresh cycles and alert thresholds
• Data visualisation selects charts according to comparison, trend, distribution or location
• Disaggregation displays equity gaps without exposing identifiable participant information
• Control charts distinguish common-cause variation from signals requiring investigation
• Accessible design applies colour contrast, plain language and alternative text
• Data stories combine annotated visuals, uncertainty statements and recommended actions

Participants can design credible evaluations that combine digital data with qualitative and quantitative methods.

• Mixed-methods designs integrate platform analytics, surveys, interviews and observation
• Quasi-experimental designs assess contribution using matching, difference-in-differences and regression discontinuity
• Digital trace data require construct validation before outcome interpretation
• Sentiment analysis findings are triangulated against interviews and contextual evidence
• Social network analysis maps information flows, influence and referral patterns
• Adaptive evaluation questions respond to implementation signals without compromising comparability
• Reproducible analysis records code, versions, decisions and data transformations

Participants can prepare a governed, feasible plan for improving digital M&E systems and methods.

• Digital M&E roadmaps sequence priorities, owners, dependencies, budget and milestones
• Stakeholder mapping assigns decision rights to communities, partners, funders and vendors
• Procurement specifications define interoperability, data ownership, support and exit requirements
• Responsible innovation reviews assess benefit, harm, inclusion and environmental costs
• Change management plans address adoption, training, resistance and feedback channels
• Implementation scorecards track readiness, delivery risks, corrective actions and learning
• Personal application plans specify a ninety-day M&E technology improvement