Exhibit 35.1 Sample size is a commercial decision
that weighs the costs of a larger sample against the benefits of greater
accuracy.
Sampling, by design, is imperfect and sample-based
findings are never 100% accurate. However, they are meaningful and consequential if the study
is well designed.
The determination of sample size involves practical and commercial
considerations that weigh the costs of a larger sample against the benefits of greater
accuracy (refer to illustration in Exhibit 35.1).
There is not much value in the information sourced from a sample unless it can be
generalised to the target population. The capacity to do so with some confidence depends on
factors associated with sample design as well as non-sampling inaccuracies.
Small unreliable samples that do not permit generalization are not meaningful
or useful. Conversely, large, overly accurate samples may be needlessly expensive. An ideal
sample is one that precisely meets specifications — it is neither over specified nor
underspecified. The determination of ideal sample size is dependent on the following factors:
- Population variability. The greater the variability the larger the
sample required to achieve the desired level of accuracy. In the case of continuous
variables such as an item’s averages sales, this variability is best reflected by the
relative standard deviation.
- Sample design. For services like retail audits, a stratified sample
design can yield a substantial reduction in sample size.
- Specified level of accuracy. Specified level of accuracy. The
standards for acceptable sampling error are set by the service provider. The higher the
required precision, the larger the sample size.
Other factors specific to the nature of the research also affect sample size.
For instance, for retail audits, larger sample of retail stores are required if products are
thinly distributed. Similarly, for usage and attitude studies, if product usage is low, a
larger sample of consumers will be required to achieve the desired level of accuracy.
It is pertinent to note that sample size is not dependent on universe size.
This may sound counterintuitive — if universe size is not a factor, why then do we need large
retail audit samples in big countries like China and India? The reason is that these markets
exhibit greater variability and lower product distribution. Furthermore, large markets often
have numerous market breakdowns, such as regions, provinces, and cities. Each market breakdown
requires its own sample size to meet the specified accuracy standards, which increases the
overall requirement.
When designing a sample, the level of data precision must be established based
on the nature of the research objectives. Industry or agency norms often serve as guidelines
for a wide range of market research programs, such as NielsenIQ’s sampling standards in retail measurement.
The level of precision is typically determined by specifying the standard
deviation (standard error) of the parameter to be estimated or by stipulating the desired
probability of achieving statistical significance for a particular estimate.
In addition to the precision levels, the calculation of sample size requires
knowledge or assumptions about the market being studied. Some of this information is sourced
from large-scale studies, such as household surveys or establishment surveys like the
retail census for retail tracking
studies. Other information, such as the proportion of the population using a brand, may be based
on conservative assumptions.
