Sharif Analytics — Quality Policy
The standard we hold ourselves to — categorized by service pillar, so evaluators can see exactly how quality is controlled in the work they’re commissioning.
Overview
Quality at Sharif Analytics is not a single service offering — it is the operating discipline behind every engagement we deliver, from research design to engineering supervision to program evaluation. This policy sets out the standard our team is held to, led by our founder’s 15+ years of practitioner experience in Lean Six Sigma and Japan-trained quality management, and translated into pillar-specific controls below so partners can see how it applies to their specific engagement.
Our Quality Principles
- Evidence Before Recommendation. No advisory output leaves our firm without being grounded in verified data, tested methodology, or field-validated findings — across Research, Engineering, and Social Development work alike.
- Independent Verification. Wherever our scope includes audit or assurance work — technical, quality, or program-level — findings are independently verified, not assumed from client-provided information alone.
- Methodological Rigor. We apply internationally recognized standards and methodologies (Lean Six Sigma, TQM, Kaizen, TPM, and recognized research/M&E frameworks), adapted appropriately to local context.
- Continuous Internal Review. Every engagement is reviewed against defined quality checkpoints before delivery — not only at project completion, but at each major milestone.
- Accountability to Client Outcomes. Our quality standard is measured by whether recommendations are actually implementable and adopted, not solely by report completion.
- Confidentiality & Data Integrity. Client and beneficiary data collected during research, M&E, or audit work is handled with strict confidentiality and data protection practices throughout every stage of an engagement.
Quality Control by Service Pillar
This is the section donor and development-organization evaluators tend to read most closely. Each pillar below lists the concrete checkpoints applied — not just the principle behind them.
Research & Consultancy — Quality Control in Research Methodology
- Study design: Sampling frameworks are statistically defensible, with sample size and design (probability, stratified, or purposive as appropriate) determined against the study’s confidence and precision requirements before fieldwork begins.
- Tool piloting: Every survey, FGD, and KII instrument is piloted and revised against the ToR objectives before full rollout — not deployed on a first draft.
- Enumerator training: Field teams are trained to a standardized protocol and checked for inter-rater consistency before independent deployment.
- Field-level QA: Digital data collection (e.g. KoBoToolbox/ODK-based tools) with real-time monitoring, GPS and timestamp validation, and back-checks on a defined percentage of completed interviews.
- Data cleaning: Range checks, logic checks, and outlier review are applied before analysis; discrepancies are traced back to source rather than silently corrected.
- Triangulation: Quantitative findings are cross-checked against qualitative evidence (FGDs, KIIs, case studies) before being reported as conclusions.
- Analysis & peer review: Statistical analysis is conducted using appropriate software (SPSS/STATA/R) by a named analyst, and findings undergo internal peer review against the ToR before the draft is released to the client.
- Ethical compliance: Informed consent, confidentiality safeguards, and (where required) ethical clearance are built into every data collection protocol.
- Traceability: Every reported finding is traceable back to raw data — no conclusion is included in a final report that cannot be shown against its source.
Engineering
- Independent technical review: Design outputs are checked by an engineer independent of the original design team before submission.
- Testing standards: Materials and workmanship are verified against recognized codes (BNBC, ASTM, AASHTO, BS) rather than accepted on contractor certification alone.
- Site supervision checkpoints: Field verification and quality inspections are scheduled at defined project milestones, not only at completion.
Social Development
- MEAL framework integrity: Indicators, data collection tools, and reporting formats follow a defined Monitoring, Evaluation, Accountability and Learning structure aligned to the project log frame.
- Safeguarding compliance: Engagements involving direct contact with beneficiaries — particularly women, children, and vulnerable groups — follow documented Prevention of Sexual Exploitation and Abuse (PSEA) and child-safeguarding protocols.
- Accountability to affected populations: Beneficiary feedback and complaint channels are built into program design, not added retroactively.
AI & Digital Innovation
- Human-reviewed outputs: AI-assisted outputs are reviewed by a qualified team member before delivery — no automated output is passed to a client without human sign-off.
- Oversight boundary: No engagement uses AI for unsupervised decision-making on matters affecting beneficiaries, funding, or reporting conclusions.
Quality (Training & Development)
- Internal consistency: The same Lean Six Sigma and TQM standards we deliver to clients through training are applied internally to our own engagement delivery — our quality practice is not separate from what we teach.
Leadership Commitment
This policy is set and reviewed by our founder, Mohammad Shariful Islam, a certified Lean Six Sigma Master Black Belt with Japan-trained quality management expertise — ensuring the standard applied to client engagements is the same standard our own leadership was trained to uphold.
Want to know how our quality standard applies to your project?
Talk to our team about the quality checkpoints built into your specific engagement.