Textual and contextual analysis of professionals’ discourses on XBRL data and information quality

• DOI: https://doi.org/10.1108/IJAIM-01-2018-0003
• Published date: 05 August 2019
• Date: 05 August 2019
• Pages: 492-511
• Author: Arif Perdana,Alastair Robb,Fiona Rohde

Textual and contextual analysis of

professionals’discourses on XBRL

data and information quality

Arif Perdana

Design and Specialised Business Cluster,

Singapore Institute of Technology, Singapore, Singapore

Alastair Robb

Business School, The University of Queensland, Brisbane, Australia, and

Fiona Rohde

TC Beirne School of Law and UQ Business School, The University of Queensland,

Brisbane, Australia

Abstract

Purpose –The purpose of this study is to gain insight intowhat aspects of eXtensible Business Reporting

Language(XBRL) data and information quality (DIQ) mostinterest professionals.

Design/methodology/approach –The authors use text analytics to examine XBRL discourses from

professionalsworking in the domain. They explore the discussionin the three largest LinkedIn XBRL groups.

Data collectioncovered the period 2010-2016.

Findings –Via the text analytics, the authors find the most appropriate XBRL DIQ dimensions. They

propose an XBRL DIQ framework containing18 relevant DIQ dimensions derived from both the accounting

and IS fields. The findings of this study are expected to help direct future XBRL research into the DIQ

dimensionsmost worthy of further empirical investigation.

Originality/value –XBRL is the internationalstandard for the digital reporting of financial, performance,

risk and complianceinformation. Although the expectations of XBRL to produce improvements in DIQ via its

applications (e.g. standard business reporting, digital data standard and interactive data visualization) are

high, they remain unclear. Thispaper contributes to better understanding of the aspects of XBRL DIQ most

relevantto professionals.

Keywords XBRL, Information quality, eXtensible Business Reporting Language, Data quality

Paper type Research paper

Introduction

The established literature advocates that eXtensible Business Reporting Language (XBRL)

be used to support a variety of policies such as transparency, accountability, reducing the

administrative burden in financial information disclosures, as well as financial institution

supervision and risk management (Liu, 2013; Perdana et al., 2015). Prior studies found that

XBRL had positive economic impacts on firms by, for example, reducing the cost of equity

capital (Hao et al.,2014), lowering the cost of debt for firms (Lai et al.,2015), and improving

forecast accuracy (Liu and O’Farrell, 2013). In the USA, Canada and Asia, XBRL is

implemented to improve accountability and transparency of financial statementsin capital

markets (Callaghan and Nehmer, 2009;Kimet al., 2012;Chen et al.,2016). In Europe, XBRL

is used to help enable the activities of banking supervision, risk management,

tax administration, datastandardization for exchanging business and financial information

IJAIM

27,3

492

Received9 January 2018

Revised11 July 2018

Accepted1 August 2018

InternationalJournal of

Accounting& Information

Management

Vol.27 No. 3, 2019

pp. 492-511

© Emerald Publishing Limited

1834-7649

DOI 10.1108/IJAIM-01-2018-0003

The current issue and full text archive of this journal is available on Emerald Insight at:

www.emeraldinsight.com/1834-7649.htm

and governance-related filings (Apostolou and Nanopoulos, 2009;Bonson, 2001;

Bonson et al., 2009a,2009b;Guillouxet al.,2013;Troshani et al.,2015). In Australia, XBRL is

used to help support the Australian Standard Business Reporting (SBR) instantiation

(Piechocki, 2010:Robb et al., 2016). XBRL can, therefore, help improve administrative,

financial and accountingprocesses and reporting for both the private and public sectors.

A substantial number of papers, ranging from conceptual to research papers, has been

published in the area of XBRL including its applications, for example, standard business

reporting (Robb et al., 2016), digital data standard (Troshani and Lymer, 2010;Guilloux

et al.,2013),and interactive data visualization (Janvrin et al.,2014;Locke et al.,2015;Perdana

et al.,2018). One of the focuses of XBRL research is to improve the data and information

quality (DIQ) of financial information (Yuan and Wang, 2009;Muthusamy et al.,2017).

Research has dealt with severalDIQ dimensions that XBRL can enhance, such as relevancy,

transparency, integrity, credibility, comparability, consistency, understandability,

reliability, automated data, and value added (Mejzlik and Istvanfyova, 2008;Yuan and

Wang, 2009;Ditter et al.,2011;Vasarhelyi et al.,2012;Muthusamy et al., 2017). In contrast,

research has also found that DIQ errors appeared in early XBRL filings. Of the 435 XBRL

filings to the US Securities and Exchange Commission (SEC) in 2010, a quarter of them

contained calculation errors (Debreceny et al., 2010). XBRL filing errors, however, have

decreased since the initial XBRL implementation in the US (Bartley et al.,2011). The

decreased error rate of XBRL filings was found to be associated with companies’

experiential learning (Du et al., 2013). The above findings indicate that addressing DIQ is

critical to the improvementof XBRL developments and implementations.

XBRL research has largely used an accounting information quality framework to

evaluate and analyze XBRL DIQ by empirically investigating dimensions such as, ease of

understanding, comparability, relevancy, reliability, timeliness, and value added (Hodge

et al.,2004;Yuan and Wang, 2009;Arnold et al.,2010;Muthusamy et al., 2017). XBRL can

also improve users’experiences when browsing and searching for information and finding

relevant materialin financial reports (Hodge et al.,2004;Arnold et al.,2010).

Given its multidimensionality, DIQ should be assessed using the attributes of

information sources that are relevant to quality (Wang and Strong, 1996). Further, DIQ is

inextricably linkedto the environment where the data and informationare used (McKeating,

1992). This linkage implies that the purposeand the context of the data and information use

necessitate different DIQ dimensions. As XBRL intersects both the accounting and

information systems (IS) fields, we argue that the applicable DIQ dimensions for XBRL

should, likewise,be derived from both the accounting and IS fields.

While prior research in XBRL has derived DIQ dimensionsfrom the accounting and the

IS literature, the research evaluateda limited number of XBRL DIQ dimensions (e.g. ease of

understanding, value added, and relevancy, reliability, understandability, timeliness and

comparability). While DIQ is one of the major concerns arising in professional discourses

(Perdana et al., 2015b), there appearsto be no specific framework in either the accounting or

IS fields for assessing XBRL DIQ. We argue that identifying relevant DIQ dimensions will

enable organizations to better undertakeDIQ assessments leading to improvements in, and

better management of, XBRL development and implementations. As noted by Wang and

Strong (1996), DIQ can be improved by understanding which aspects of it most interest its

users. Therefore, we contribute by providing a relevant multidimensional DIQ framework for

XBRL by incorporating DIQ dimensions from both the accounting and IS fields. This proposed

framework can be empirically investigated in future research. In line with the preceding

research motivations, we pose the following research questions to help guide our study: What

Textual and

contextual

analysis

493