Sampling unit
The minimum unit of observation in a social impact assessment's sample design — a social unit such as a family, or a geographical unit such as a village — the basic building block behind every sample figure.
In plain language
Before an assessor can pick a sample of "5%" or "8%" of a population, someone has to decide: 5% of what, exactly? Of people? Of families? Of villages? That basic building block is the sampling unit.
The workbook's definition: the sampling unit is "the minimum unit for observation, that acts as a building block of the data." It can be a social unit — a family, a school, a place of worship — or a geographical unit — a state, district, block, or village.
Choosing the right sampling unit comes before choosing a sample size. If beneficiaries live in households, the natural sampling unit might be the family, not the individual — counting every family member separately, when they share one set of circumstances, can distort the picture.
How it works
The sampling unit is the second of four steps the workbook sets out for determining a sample (Chapter 5, section 5.10, "Sample Size"):
- Type of Universe — defining the full population (finite or infinite in number)
- Sampling Unit — the minimum unit for observation: a social unit (family, school, religious place) or a geographical unit (state, district, block, village)
- Sample Size — how many units, from the universe, are actually selected; depends on time, cost and facility
- Sampling Methodology — influenced by the evaluation's objectives, resources, time and the nature of the problem; calculating a statistically significant sample size needs a defined confidence interval and margin of error
The choice of sampling unit shapes everything after it. If the sampling unit is a village rather than a household, the same 5% sample figure covers a very different, and much smaller, number of people actually interviewed.
A worked example
Drawing on the workbook's own case study figures.
In the IWMP watershed project, spanning 18 villages with a combined population of 25,000, the assessor takes a 5% sample of direct beneficiaries — farmers, local business owners and community members.
Here the sampling unit is effectively the individual beneficiary within the village population, not the village itself — a 5% sample of 25,000 people means roughly 1,250 people are approached, spread across all 18 villages, rather than 5% of the 18 villages.
Had the assessor instead chosen "village" as the sampling unit, a 5% sample would round to essentially zero villages — showing why picking the right sampling unit, before setting a sample-size percentage, changes what that percentage actually means on the ground.
Why NISM asks about it
Chapter 5 (Social Impact Assessment and Social Impact Assessors), section 5.10, defines the sampling unit as the second of four steps in determining a sample, alongside universe, sample size and sampling methodology. Chapter 8's case studies use sample percentages (5% in the IWMP case, 8% in the E-Diksha case) that only make sense once the sampling unit is fixed. Expect a question asking what a sampling unit is.
Common exam traps
- A sampling unit is not the same as sample size. The unit is what is being counted; the size is how many of those units are selected.
- A sampling unit can be a social unit or a geographical unit — a question may test whether a candidate can tell the two categories apart.
- Choosing the sampling unit comes before choosing sample size and methodology in the workbook's own four-step order.
- A sample percentage (5%, 8%) is meaningless without knowing the sampling unit it is a percentage of.
Where this is taught
Free preparation for NISM Series XXIIIRelated terms
- Desk reviewThe first step of a social impact assessor's fieldwork: examining an organisation's existing documents — reports, policies, photographs, prior data — before any physical inspection or interview.
- Social Impact AssessmentA systematic evaluation of the social, ethical, cultural and environmental consequences of a project or organisation — positive and negative, intended and unintended — taking an "outside in" view.
- Social Impact AssessorAn individual registered with an SRO (under ICAI or another SEBI-specified agency) who has passed the NISM certification and holds a valid certificate; performs independent verification of impact reports.
- Quantitative approachOne of three approaches to Social Impact Assessment, using surveys and statistical analysis to process large volumes of data into results stakeholders can quantify and compare.
- SAS 500The thematic standard for environmental sustainability, climate change and conservation, illustrated by AB Responsible Services' watershed programme across 18 villages in Maharashtra.