Identifying XBRL’s data information quality dimensions using text mining and topic analysis

The expectations for improvements to Data and Information Quality (DIQ) provided by XBRL are high, however, the applicable DIQ dimensions of XBRL remain unclear. To help gain insights into the relevant DIQ dimensions we explore professional perspectives relative to XBRL’s DIQ. We used professional discussions in social media, particularly in LinkedIn groups, to help obtain such perspectives. Prior research in XBRL has derived the DIQ dimensions largely from information systems (IS) and accounting literature. The IS and accounting research, however, only evaluated a limited number of XBRL’s DIQ dimensions (e.g., ease of understanding, value added, and relevancy, reliability, understandability, timeliness and comparability). The IS and accounting research fields, however, have yet to formulate a framework that assesses XBRL’s DIQ. This paper explores the discussion taking place on LinkedIn to seek insights into which DIQ dimensions most interest XBRL users. Text mining and topic analysis using sample data from the three largest LinkedIn XBRL groups were conducted to uncover the most relevant DIQ dimensions of XBRL. The findings of this study are expected to help direct future research into the DIQ dimensions of XBRL that should be empirically investigated.