Research Design and Methods Training Course

Participants can construct and test a Theory of Change for a proposed study.

• Constructing a Theory of Change from activities to long-term outcomes
• Results chains separate inputs, activities, outputs, outcomes and impacts
• Testing causal assumptions, external factors and plausible alternative explanations
• Outcome pathways specify actor behaviour, service quality and institutional conditions
• Selecting indicators that represent each link in the causal pathway
• Indicator reference sheets define numerator, denominator, source and disaggregation

Participants can translate abstract programme concepts into measurable study variables.

• Constructs distinguish latent concepts from directly observable programme variables
• Operational definitions specify exactly what each measure includes and excludes
• Conceptual frameworks identify expected relationships among variables and outcomes
• Turning broad questions into measurable primary and secondary study questions
• Exposure, outcome, mediator, moderator and confounder variables clearly distinguished
• Developing measurement matrices linking questions, variables, methods and respondents

Participants can select study designs suited to questions, resources and ethical constraints.

• Cross-sectional designs estimate conditions, associations and population characteristics at one point
• Longitudinal designs track change through repeated measures with defined follow-up periods
• Case-control studies compare prior exposures between outcome-defined participant groups
• Cohort studies estimate incidence and temporal relationships among exposed populations
• Case study designs examine implementation mechanisms within bounded programme settings
• Design selection balances decision value, feasibility, ethics, time and available data

Participants can assess options for estimating programme effects and causal attribution.

• Randomized controlled trials allocate comparable units through documented random assignment
• Cluster randomization addresses interventions delivered through communities, schools or facilities
• Difference-in-differences compares outcome trends between intervention and comparison groups
• Regression discontinuity exploits eligibility thresholds with precisely measured assignment variables
• Propensity score matching creates comparison groups from observed baseline characteristics
• Contribution analysis assesses causal contribution where counterfactual designs remain infeasible

Participants can select qualitative approaches that investigate experience, process and meaning.

• Phenomenological inquiry investigates participants' lived experiences of services and change
• Grounded theory develops explanatory concepts from systematically coded field evidence
• Ethnographic observation documents social practices, interactions and contextual programme conditions
• Participatory rural appraisal uses community-led mapping, ranking and seasonal calendars
• Most Significant Change captures stakeholder accounts of meaningful programme effects
• Qualitative case selection follows relevance, variation, information richness and safeguarding

Participants can design probability samples and calculate defensible sample sizes.

• Sampling frames define eligible units, coverage gaps and duplicate-record risks
• Simple random sampling assigns equal selection probability to every eligible unit
• Systematic sampling selects units using intervals after a randomized starting point
• Stratified sampling improves representation of priority groups and analytical subgroups
• Cluster sampling reduces field costs when households are geographically dispersed
• Sample size calculations incorporate prevalence, precision, confidence and design effect

Participants can plan qualitative sampling and recruitment with transparent selection criteria.

• Purposive sampling recruits participants with direct knowledge of study phenomena
• Maximum variation sampling captures contrasting experiences, contexts and programme outcomes
• Criterion sampling applies explicit eligibility conditions before participant recruitment
• Snowball sampling reaches hidden populations through participant referral pathways
• Thematic saturation monitors whether additional interviews generate substantively new findings
• Recruitment plans document gatekeepers, consent procedures, incentives and refusal handling

Participants can draft survey instruments that produce interpretable and reliable data.

• Question wording uses one concept, clear recall periods and neutral language
• Closed response categories remain mutually exclusive and collectively exhaustive
• Likert scales specify balanced response options and consistent directional anchors
• Skip logic routes respondents past questions that do not apply
• Sensitive questions use privacy safeguards, indirect wording and voluntary response options
• Cognitive interviewing identifies misunderstanding, recall difficulty and problematic response processes

Participants can develop interview, focus group and observation instruments for field use.

• Semi-structured interview guides sequence opening, probing and closing questions logically
• Focus group guides distinguish shared norms from individual experiences and dissent
• Probes elicit examples, sequences, reasons and exceptions without leading respondents
• Observation protocols record setting, behaviour, interactions and researcher reflections separately
• Document review matrices extract evidence against predefined questions and criteria
• Piloting instruments tests duration, terminology, flow, translation and recording arrangements

Participants can identify and manage ethical, safeguarding and privacy risks in study plans.

• Informed consent communicates purpose, procedures, risks, benefits and withdrawal rights
• Assent and guardian permission address research involving children and adolescents
• Do-no-harm assessments identify potential stigma, retaliation, distress and exclusion
• Safeguarding referral protocols specify response routes for disclosed protection concerns
• Data minimization limits collection to information necessary for stated study purposes
• Pseudonymization separates personal identifiers from analytical research datasets

Participants can prepare fieldwork plans that protect data quality under operational constraints.

• Fieldwork plans assign roles, locations, schedules, equipment and escalation procedures
• Enumerator training covers consent, instrument administration, neutrality and safeguarding actions
• Standard operating procedures document consistent collection, transfer and storage steps
• Back-checks verify selected interviews through independent respondent follow-up procedures
• Daily quality checks identify missing values, outliers, duplicates and skip errors
• Translation and back-translation preserve item meaning across required study languages

Participants can prepare quantitative datasets and analysis plans aligned with research questions.

• Codebooks document variable names, labels, formats, values and missing-data codes
• Data cleaning rules address duplicates, impossible values, outliers and inconsistent responses
• Descriptive statistics summarize distributions using counts, proportions, means and medians
• Cross-tabulations compare outcomes by sex, age, disability and location
• Confidence intervals communicate sampling uncertainty around estimated population values
• Analysis plans pre-specify tables, subgroup comparisons, models and decision rules

Participants can organize, code and interpret qualitative evidence systematically.

• Transcription standards preserve spoken content, pauses, anonymization and translation decisions
• Deductive codes derive from research questions, frameworks and predefined constructs
• Inductive coding captures unanticipated patterns emerging from participant accounts
• Codebooks define labels, inclusion criteria, exclusion criteria and example excerpts
• Thematic analysis compares patterns across participant groups, sites and programme stages
• Analytic memos document interpretation decisions, rival explanations and researcher reflexivity

Participants can combine quantitative and qualitative evidence into coherent study conclusions.

• Convergent designs collect quantitative and qualitative evidence during similar study phases
• Explanatory sequential designs use interviews to explain quantitative patterns and outliers
• Exploratory sequential designs develop survey measures from preliminary qualitative findings
• Joint displays compare numerical results with themes, quotations and interpretation
• Triangulation examines convergence, complementarity, disagreement and gaps between evidence streams
• Integrated conclusions distinguish corroborated findings from unresolved mixed-methods contradictions

Participants can diagnose quality threats and specify appropriate mitigation measures.

• Construct validity assesses whether measures represent their intended underlying concept
• Internal validity examines competing explanations for observed programme outcome differences
• External validity considers transferability beyond sampled people, places and conditions
• Inter-rater reliability tests consistency between observers, coders or assessors
• Selection bias arises when included participants differ systematically from target populations
• Recall and social desirability bias require design, wording and administration controls

Participants can judge whether evidence supports causal claims and programme recommendations.

• Counterfactual reasoning asks what outcomes would occur without programme exposure
• Confounding occurs when third variables influence both exposure and measured outcomes
• Temporal precedence requires causes to occur before proposed effects are observed
• Dose-response patterns strengthen causal interpretation when exposure intensity varies meaningfully
• Rival explanations include concurrent programmes, seasonality, shocks and measurement changes
• Evidence appraisal separates association, contribution, attribution and plausible mechanism claims

Participants can communicate findings, limitations and recommendations for decision-making.

• Research reports distinguish methods, findings, interpretation, limitations and recommendations clearly
• Executive summaries present decision implications before technical methodological detail
• Tables and charts match data type, comparison purpose and audience needs
• Disaggregated reporting identifies differences without exposing small or identifiable groups
• Limitation statements explain likely direction, scale and implications of bias
• Management response plans assign actions, owners, deadlines and follow-up evidence

Participants can prepare and assess research commissions with clear methodological requirements.

• Terms of reference specify decisions, questions, scope, methods and deliverables
• Evaluation questions rank priority evidence needs when resources cannot cover everything
• Technical proposals assess design coherence, sampling adequacy and ethical protections
• Inception reports confirm methodology, workplan, instruments and analysis approach
• Quality review checkpoints examine protocols, tools, preliminary findings and draft reports
• Data ownership clauses define access, retention, transfer and publication responsibilities

Participants can build, review and present a complete research protocol for a programme decision.

• Protocol purpose defines the decision, users, questions and intended evidence use
• Theory of Change identifies outcomes, assumptions and priority evidence gaps
• Design justification compares feasible alternatives against causal and ethical requirements
• Sampling plan specifies population, frame, selection method, size and recruitment
• Data collection and analysis plans align instruments with each research question
• Peer review tests feasibility, bias controls, safeguarding and reporting usefulness