This article critically examines the historical
reliance on sampling in survey research and argues for a renewed consideration
of census-based approaches in the digital era. Traditional benchmarks have
guided generations of researchers in determining the minimum sample sizes
required. While these conventions remain influential, the growing availability
of electronic sampling frames and administrative datasets increasingly makes
census studies methodologically superior, as well as practical and
cost-effective. The rise of Big Data further blurs the distinction between
sampling and census research, enabling analyses of entire digital populations
while introducing new methodological and ethical challenges related to
representativeness, data quality, and privacy. The paper examines the economic,
practical, statistical, and ethical benefits of census research, including
higher response rates, reduced sampling error, the ease of post-stratification,
and increased inclusivity. Attention is also given to the persistent challenge
of non-response bias, with evidence suggesting that response rates, whether for
samples or censuses, remain broadly comparable. Recommendations emphasise the
adoption of census strategies where feasible, supported by digital
infrastructure. Future research should investigate response dynamics and
consider the ethical benefits associated with census approaches. Collectively,
this analysis calls for adopting census designs and rethinking survey
methodology, taking into account contemporary technological advancements.
