NGO data quality audit checklist: what to review before donor reporting
A practical NGO data quality audit checklist for M&E teams reviewing KoboToolbox, ODK, Excel or CSV data before analysis and donor reporting.
Read the guide →SPSS · R · BIOSTATISTICS · DATA QUALITY
Practical guidance for research that must survive real questions.
Clear, inspectable workflows for researchers, public-health teams and NGOs—from proposal methodology and sample size to cleaning, analysis and reporting.
BUYER & PRACTITIONER LIBRARY
Start with the problem in front of your team.
Each guide is designed to help you diagnose the next step before you book support.
KoboToolbox and ODK data quality monitoring: a practical NGO workflow
How NGO and M&E teams can monitor KoboToolbox and ODK survey data safely—from form controls and live sync to review queues, cleaning logs and reporting-ready exports.
Read the guide →How to prepare Kobo, ODK or Excel survey data for analysis
A step-by-step workflow for preparing KoboToolbox, ODK, Excel or CSV survey data for analysis without losing raw records, labels, cleaning decisions or important limitations.
Read the guide →SPSS data analysis consultant online: what credible support should include
A practical buyer's guide to remote SPSS data analysis consulting: scope, data cleaning, syntax, statistical methods, deliverables, privacy and red flags.
Read the guide →R statistical consulting for public health research: a defensible workflow
How remote R statistical consulting should handle public-health questions, reproducible code, model checks, interpretation, privacy and handover.
Read the guide →Sample size calculation for a cross-sectional study: formula, assumptions and example
Learn how to calculate sample size for a single-proportion cross-sectional study using confidence, precision, expected prevalence, finite population, design effect and non-response.
Read the guide →How to prepare data for a remote statistician without losing context or privacy
A practical checklist for preparing a question, codebook, de-identified dataset and analysis context before working with a remote statistician.
Read the guide →Survey data cleaning checklist for SPSS and Excel
A practical survey data cleaning checklist covering duplicates, coding, missingness, ranges, skip logic, dates, derived variables, logs and analysis readiness in SPSS or Excel.
Read the guide →Research proposal methodology review service: what gets checked before data collection
What a credible research proposal methodology review should assess: alignment, study design, sampling, variables, instruments, analysis, ethics, feasibility and revision notes.
Read the guide →Turn a data management plan into project actions
Make an inventory with one row per source: file type, row unit, collector, approximate volume, owner and permitted use. Separate a public reference table from identifiable interview records. They may need different storage, access and sharing rules.
Read the guide →Build a research data provenance and version manifest
A checksum helps identify the exact file used. It does not establish who originally created it, whether reuse is permitted or whether its values are correct. Record provenance uncertainty beside the identifier so it cannot disappear from the final report.
Read the guide →A worksheet for independent statistical QA
Freeze the input and output versions before QA. A reviewer cannot reliably sign off a moving collection of files. Identify the objectives, analysis plan, transformations and draft report included in the candidate, plus any unavailable evidence.
Read the guide →What belongs in a research Evidence Pack?
A programme manager may need a decision summary while an analyst needs code and diagnostics. Agree both layers before writing the report. Do not assume a polished executive summary replaces enough technical detail to inspect the result.
Read the guide →Questionnaire validation: questions to ask before alpha
Define what is measured, for whom and for which use. A scale adapted for a new population or language needs more than a copied reliability coefficient. Record permission to reuse items and the reasons for additions, removals or translations.
Read the guide →Keep a decision log for missing-data analysis
Separate structurally skipped values, unknown responses, refusal and invalid entries. A field that was never asked is different from a response that was lost. Review the instrument and eligible population before interpreting a missing-cell count.
Read the guide →Review field-quality signals without unsupported fraud claims
Retain the form version, collection window, repeat relationships and choice lists with the export. A field renamed halfway through collection can appear missing even when its information exists under an earlier name.
Read the guide →Map research objectives to an analysis plan
Write the target population, primary objective and quantity you want to estimate. “Find significance” is not an objective. Decide whether the question is descriptive, associational, predictive or intended to support causal interpretation.
Read the guide →NEED A PROJECT-SPECIFIC ANSWER?
Book a 20-minute demo ↗Bring the form, dataset or reporting risk.
Book a 20-minute demo. We will identify whether a dataset audit, field operations pilot or managed support is the right next step.