User-hostile data portals, including those of governmental agencies, have hindered researchers and data journalists from finding and using the right datasets to produce impactful works, a new study reveals.
As non-expert data users, researchers working in interdisciplinary areas and data journalists often face technical barriers when doing data search for their works, according to article “Truth in a sea of data” published in Information, Communication & Society journal in late November.
The study found that there were three main problems encountered by researchers and data journalists when exploring various data repositories, government websites, and/or commercial data hubs: searchability, interactivity, and usability.
In terms of searchability, non-expert users often had a difficult time finding relevant datasets due to the “disorderly nature” of the websites. “It was evident that there was a mismatch between users’ thought processes for formulating queries and the queries that search engines required to produce desirable results,” the study states.
Many data sites also offered few interactive features, preventing users from filtering data as well as communicating with data producers or fellow users to clarify a dataset. Moreover, most of the study participants reported difficulties in extracting data from various websites, which were not designed with non-experts in mind.
“Sometimes [the data managers] will have a table, print it, and they’ll scan it, and then they just put that in a PDF, so that the whole thing is an image,” said one study participant.