—–Prof Udaya S Mishra and Dr Umakanta Sahoo
Statistics today is all-encompassing, given the digital access to information of every kind. There remains no domain beyond the purview of statistics that literally makes a living based on statistics – be it regarding weather, prices, disasters, and several other vital features. In fact, every decision, beginning with what to buy, where to invest, when to initiate an activity of what to choose depends on the statistics available on our mobile handset.
In this kind of environment, it’s extremely important to differentiate between reliable and credible statistics from that of unreliable and unscientifically derived statistics. In this environment of big data, statistics does not remain limited to a science of compilation, computation and interpretation of information in different forms, but the conversion of a wide range of information (not necessarily numeric) into a form that can be analysed for a wider understanding of social and economic transformation at large.
Statistics and the state of affairs
To comment on any state of affairs, there is a reference to statistics. In the past, we had branches like medical statistics, agricultural statistics, consumption statistics, and, at a broader level, macro and micro statistics that were distinguished in terms of aggregates and individual unit-based statistics. But now, the current state of affairs from varied lenses requires statistics to defend a viewpoint. If someone observes that a country is politically unstable, it needs to be supported with statistics to make such an observation. Similarly, to make an observation regarding socio-economic fragility, economic prosperity to social harmony needs statistics in terms of relevant indicators.
This has led to the advent of conceptualising phenomena, perceptions and behavioural domains into numeric or scaled assessments and its valuation. Every approach of this kind is initially triggered by comparison followed by valuation, and finally, their interpretation.
Credibility deficit in this incredible expansive domain of statistics
In times of inadequate and insufficient data, the little that was available to us was credible, carefully computed and never challenged. But with the progress made on computational fronts along with advances in digitalisation, information gets generated from varied sources, giving rise to confronting and conflicting figures raising the question of credibility. Plenty has its problem and aspiration for all kinds of statistics, with micro levels of disaggregation intensifying this concern further. Owing to digitalisation, a whole range of administrative statistics gets generated, which remains underutilised due to the valid reasons of them being denominator-free statistics.
At the same time, numerous surveys being conducted to generate periodic statistics on a range of phenomena have their own folly regarding their comparability and conflicting mismatch across various surveys. Temporal and cross-sectional comparability of any statistic that is generated requires a definitional compliance as regards the phenomenon. This, too, is often violated, making comparative features of statistics compromised at varied levels.
Providers of statistics need to educate the users
When information is in the public domain in the form of statistics that is universally accessible in a digital environment, its responsible reading remains the key, given its misinterpretation more than its right interpretation. Every statistic that is made available with ease to anyone and everyone is not without its conceptualisation, measurement acumen and disciplinary orientation. Such conceptualisation undoubtedly varies between varying domains like educational statistics, agricultural statistics, health statistics, medical statistics, and many others.
What is frequently observed is that there is not only indiscriminate use of these statistics for interpretation in making observations that are quite sensitive, but these are also often misinterpreted beyond proportion to generate unnecessary controversies. The available statistics, however simple, may be like rates and proportions to that of domain-specific measures and indicators. Its responsible use and interpretation remain fundamental to the claim of evidence-based valuation based on statistics.
Wisdom in the time of plenty
The multiple and widespread features of statistics not only pose a problem of reconciliation but also lead to greater misuse rather than use of the same. The use of statistics may be assumed as simplistic, but an efficient use of statistics involves an understanding of its formulation and generation along with the purpose behind the same. On many occasions, people make reference to statistics to drive home a point of argument that is truly unrealistic. Say, for instance, there is a growth in the aviation sector reflected in the quantum of air traffic, but it does not necessarily imply that air travel has become accessible to one and all.
Similarly, varying forms of deprivation gauged with statistics need to consider evolving changes. Housing deprivation yesterday was referred to as having no roof over the head, which today is to be referred to as having a house with all basic amenities. The essential dimension, therefore, is the scientific and wiser use of statistics in inferring the state of affairs and making opinions on certain phenomena.
The present statistical ecosystem is evolving at the interface of wilder databases and computerised state-of-the-art aggregation. The entire big data platform is, in many ways, the future of the unfolding statistical ecosystem. While such a platform offers a wider canvass, a cautious and responsible exploration of the same remains a challenge on one hand and an opportunity on the other to generate new frontiers of data.
