Sampling error and public decisions: when the numbers cannot be trusted
Why a three-point gap between regions may be statistically meaningless, and how to check before the figure reaches a briefing.
Research organisation · Astana
Scicom LLP conducts scientific research and experimental development in the social sciences and humanities. We turn empirical data into reasoned conclusions that withstand academic scrutiny.
Home
Figures confirmed by reporting on completed contracts and research assignments.
Scicom LLP was registered on 18 June 2025
Including two holders of doctoral and candidate degrees
Surveys, monitoring, methodology and analysis of social processes
Cases are published in generalised form under confidentiality terms
Practice
Empirical study of social processes: from methodology design and fieldwork through to an analytical report fit for a management decision.
The research cycle
Every assignment follows the same research cycle — from formalising the question to defending the findings before the client.
We refine the research question, its boundaries and the deliverable format. The terms of reference record exactly what will be measured and what will count as a proven answer.
We develop the sample, instruments and data collection procedures. The methodology is agreed with the client before fieldwork starts.
Fieldwork with two levels of quality control: ongoing checks by the contractor and independent re-verification of a control share.
Statistical processing, hypothesis testing and a statement of limitations. Findings are presented separately from the researcher's interpretation.
Presentation to the client, questions answered and revisions made where required. The report is handed over together with the primary data set and methodology.
Analytics
Papers, reviews and reports by experts of the Scicom Analytical Centre.
Why a three-point gap between regions may be statistically meaningless, and how to check before the figure reaches a briefing.
Quantitative indicators do not explain why a programme failed. Where a qualitative stage is needed and how to build it in correctly.
Identically named indicators are calculated differently across organisations. What to check before analysis begins.
Portfolio
A selection of delivered research assignments and expert review cases.
Request an expert opinion
Outline your task in general terms — an expert will contact you within one business day to clarify the subject, timeline and format of the deliverable. All enquiries are handled confidentially.
Government and quasi-public sector
Research institutes
International organisations