Research Design, ODK Data Collection, GIS Mapping and Data Analysis Training Course

Participants can select a study design and prepare an appropriate sampling approach.

• Cross-sectional, cohort, case-control and panel designs matched to questions
• Experimental and quasi-experimental designs distinguish attribution from contribution claims
• Difference-in-differences, matching and regression discontinuity assumptions examined before selection
• Sampling frames define target populations, units and coverage errors
• Probability samples use simple random, systematic, stratified and cluster methods
• Purposive, maximum-variation and snowball sampling support qualitative inquiry
• Sample-size calculations use confidence, precision, design effect and nonresponse

Participants can define measurable indicators and plan valid data collection instruments.

• Indicator reference sheets specify definitions, numerators, denominators and disaggregation
• SMART criteria test whether indicators are useful and measurable
• Operational definitions reduce variation between enumerator interpretations
• Likert scales, ranking questions and semantic differentials measure perceptions
• Question wording avoids double-barrelled, leading and socially desirable responses
• Translation and back-translation protect meaning across language versions
• Pilot protocols test comprehension, timing, sequencing and respondent burden

Participants can build validated mobile data collection forms using XLSForm.

• XLSForm survey architecture uses survey, choices and settings worksheets
• Question types, constraints, relevance and required fields prevent invalid entries
• Repeat groups capture household members, facilities, commodities and service episodes
• Choice filters and cascading selects reduce enumerator search time
• Calculations, defaults and metadata record derived values and audit fields
• Form logic tested with ODK Collect before server deployment
• Multilingual labels and media attachments support accessible respondent interactions

Participants can configure an ODK deployment and manage controlled digital data collection.

• ODK Central projects, user roles and server permissions configured
• Android device setup secures ODK Collect, storage and synchronization
• Enumerator briefing covers consent scripts, neutral probing and referral procedures
• Pilot results drive wording revisions, timing estimates and skip-logic corrections
• Daily synchronization monitors submissions, edits, duplicates and late records
• Fieldwork dashboards track progress against clusters, quotas and targets
• Incident logs document safeguarding concerns, refusals and device losses

Participants can prepare and manage geographic data for programme monitoring and evaluation.

• Coordinate reference systems distinguish geographic coordinates from projected measurements
• Administrative boundaries, points, lines and polygons represented as vector layers
• Raster surfaces represent elevation, rainfall, land cover and satellite imagery
• QGIS project structure manages layers, symbology, labels and metadata
• Geocoding joins survey locations to settlements and service facilities
• GPS accuracy, displacement and confidentiality risks assessed before mapping

Participants can plan qualitative data collection that produces credible and ethically gathered evidence.

• Interview guides align open questions with evaluation criteria and indicators
• Focus group protocols manage participation, power differences and sensitive topics
• Observation checklists capture facilities, behaviours, processes and contextual conditions
• Key informant selection documents expertise, positionality and potential bias
• Consent, assent and safeguarding procedures protect vulnerable research participants
• Audio, notes and transcripts follow secure file-naming conventions
• Reflexive memos record researcher influence, emerging themes and decisions

Participants can structure research data for secure, traceable and efficient analysis.

• ODK Central exports produce CSV files, media archives and submission metadata
• Data dictionaries define variables, labels, codes, types and allowable values
• Unique identifiers link respondents, households, sites, visits and repeat records
• Tidy data principles separate observations, variables and analytical units
• Missingness codes distinguish refusal, not applicable, unknown and skipped responses
• Version-controlled folders preserve raw, cleaned, analysis and output files

Participants can identify, investigate and document data quality issues before analysis.

• Routine data quality assessment examines accuracy, completeness, timeliness and integrity
• Range checks, duplicates and logical consistency rules identify record-level errors
• Spot checks compare source documents, submissions and reported aggregates
• Outlier investigation distinguishes valid extremes from entry and transmission errors
• Audit trails document corrections, exclusions, recoding decisions and approvers
• Data quality dashboards prioritize exceptions for supervisor follow-up

Participants can analyse survey data and generate repeatable outputs using R.

• RStudio projects separate scripts, inputs, outputs and package dependencies
• Import CSV and Excel files using reproducible R scripts
• Data wrangling with dplyr filters, recodes, joins and reshapes records
• Descriptive statistics report frequencies, central tendency, dispersion and disaggregation
• Confidence intervals and hypothesis tests support defensible group comparisons
• ggplot2 charts communicate distributions, trends, comparisons and uncertainty
• R Markdown produces repeatable tables, charts and analytical reports

Participants can use Python to automate routine data preparation and monitoring calculations.

• Jupyter notebooks combine Python code, narrative, outputs and review notes
• pandas dataframes import, inspect, clean and merge survey datasets
• Functions automate recurring validation, recoding and indicator calculations
• Regular expressions standardize inconsistent text, dates and identifier formats
• matplotlib and seaborn create diagnostic plots for quality review
• Script parameters reproduce monthly monitoring updates from new extracts

Participants can code qualitative sources and develop traceable thematic findings using NVivo.

• NVivo project structure links sources, cases, classifications and coding frameworks
• Deductive codes derive from questions, theory of change and criteria
• Inductive coding captures unexpected themes, language and respondent perspectives
• Codebook definitions specify inclusion, exclusion and illustrative evidence
• Matrix coding queries compare themes by gender, location and stakeholder
• Inter-coder agreement procedures resolve divergent interpretations and code application
• Framework matrices synthesize cases, themes, quotations and analytical memos

Participants can analyse spatial patterns and produce responsible maps for programme decisions.

• Point-in-polygon joins assign survey records to administrative areas
• Buffer analysis estimates populations within walking distance of services
• Choropleth classification uses quantiles, natural breaks and equal intervals
• Kernel density and hotspot methods reveal clustering patterns cautiously
• Spatial joins combine facility, population, hazard and survey datasets
• Map layouts apply scale bars, north arrows, sources and disclaimers
• Small-area disclosure controls protect respondents in published maps

Participants can integrate quantitative, qualitative and contextual evidence into defensible conclusions.

• Joint displays align quantitative results, qualitative themes and evaluation questions
• Triangulation compares convergence, complementarity, disagreement and unanswered evidence
• Contribution analysis tests causal claims against alternative explanations
• Equity analysis disaggregates findings by sex, age, disability and location
• Outcome harvesting documents observable changes and verifies contribution evidence
• Most significant change stories complement indicator trends with participant meaning

Participants can turn analytical outputs into proportionate findings and assigned actions.

• Effect sizes distinguish practical importance from statistical significance
• Confidence, uncertainty and limitations stated without overstating causal claims
• Subgroup findings assessed for sample adequacy, bias and disclosure risk
• Data visualization selection matches message, audience and decision required
• Results chains connect findings to corrective actions and accountable owners
• Recommendation matrices specify action, rationale, priority, owner and deadline

Participants can produce reports, dashboards and briefings that support timely management decisions.

• Evaluation report structure follows purpose, methods, findings, conclusions and recommendations
• Executive summaries state decision implications before technical detail
• Indicator dashboards use definitions, targets, trends and disaggregation consistently
• GIS map products communicate coverage gaps, access barriers and geographic inequities
• Data stories combine charts, quotations, maps and decision prompts
• Annexes document sampling, instruments, codebooks and analytical limitations
• Briefing presentations handle challenge questions about methods and evidence

Participants can apply practical controls for sensitive research and monitoring data.

• Data protection impact assessments identify collection, storage, sharing and retention risks
• Data minimization limits personal information to justified research purposes
• Role-based access controls govern identifiers, transcripts, locations and exports
• De-identification removes direct identifiers and assesses re-identification pathways
• Retention schedules define archival, destruction and handover decisions
• Humanitarian data responsibility principles guide sensitive population data handling
• GDPR principles of lawfulness, purpose limitation and accountability applied

Participants can manage fieldwork quality, enumerator performance and operational risks.

• Fieldwork plans sequence recruitment, travel, supervision, synchronization and contingencies
• Enumerator performance metrics combine productivity, quality flags and observation results
• Back-checks, accompanied interviews and call-backs verify data collection quality
• Adaptive sampling logs record replacements, access constraints and protocol deviations
• Security protocols address device theft, connectivity loss and emergency escalation
• Daily review meetings assign corrections, retraining and field decisions
• Closeout procedures reconcile devices, records, consent forms and expenses

Participants can connect research outputs to ongoing monitoring and evaluation management processes.

• Indicator reference sheets specify definitions, formulas, sources, frequency and ownership
• Baseline, midline and endline plans maintain comparable measurement methods
• Adaptive management cycles convert monitoring evidence into documented decisions
• Donor logframes align outputs, outcomes, indicators, assumptions and reporting dates
• Evaluation terms of reference define scope, questions, governance and deliverables
• Data use workshops prioritize findings by urgency, feasibility and expected benefit

Participants can build, quality-assure and present an end-to-end mixed-methods monitoring study using ODK, GIS and analytical outputs.

• Research protocol defines decision question, design, sampling and ethical safeguards
• XLSForm captures indicators, consent, constraints, repeats and location data
• ODK submissions are exported, cleaned and documented with reproducible scripts
• QGIS map visualizes geographic patterns, service access or coverage gaps
• NVivo coding or quantitative analysis produces traceable findings
• Triangulated conclusions distinguish evidence, interpretation, limitation and recommendation
• Peer review tests methods, data quality, maps and decision usefulness
• Presentation defends findings, recommendations and implementation priorities to decision-makers