|'Target' by Nina Matthews |
under a CC license
Since our world in becoming increasingly driven by data-fed decisions (in theory, at least), it may be useful to check, from time-to-time, the route from raw data acquisition, to data aggregation, processing and analysis, to the data-driven decisions, their implementation and their verification.
I don't know if the overall data production has increased, world-wide. I would suspect it has. An increased number of tasks are accompanied by data recording for the shake of quality monitoring, performance evaluation, security, etc. That is, not taking into consideration activities that aim at data production itself, such as research activities, meta-analysis studies, etc.
Access to data is another issue. Personally, I feel access to data has improved due to (i) more discrete data points/ sets being publicly available and indexed in an convenient way (e.g., search engines), (ii) more data being available in an organised, searchable form (e.g., databases) and (iii) more information on how to get close to the various data sets/ databases in publicly available (e.g., in publications, the press, search engines, etc.). What is hard for me to know is whether the data seemingly available for each instance of need is truly comprehensive or, at least, reasonably representative of the total data truly available. Further to that, it is a priori unclear, whether the data available is representative of the population under study or suitable to address the problem in question. Proper definition of the problem and careful examination of the data can provide answers to many of the questions above, of course, but in this case I'm just voicing the general concerns.
Data analysis also seems to be easier today. Sometimes, too easy. Sure, there are plenty of people that know how to use the right tools on data but, since such tools are becoming increasingly user-friendly and accessible, I suspect that not all that use them really know what they are doing (on their defense, some problems are difficult to define). And if these people are involved in decision-making then things can become questionable (or funny). Ah, and despite all the hype on big data, people still need to do some thinking!
The last bit is implementing actions based on the results of data analysis, taking into account ethical considerations and ensuring that those actions (i) actually correspond to what the results have identified and (ii) that they work as intended. Politics and policy making is not a purely science-based territory but, rather, an arena where scientific results, gut impulses, interests from various parties and plain luck - to mention a few factors only - compete to affect decisions. I'm not rushing to label that 'wrong'. After all policy making has various expectations to meet and, be it as it may, it has been an integral part of our civilisation.
I strongly feel that we should continue studying our policy making processes (there are various approaches to that), we should further improve our quality control mechanisms and develop the courage to critically review policies and measures in a transparent way at a regular fashion.