Entering data correctly: An empirical evaluation of the theory of planned behaviour in the context of manual data acquisition

https://doi.org/10.1016/j.ress.2018.05.009Get rights and content

Highlights

  • The theory of planned behaviour is tested in the context of data entry.

  • Two seldomly measured constructs are included: behavioural control and behaviour.

  • Behavioural control had a stronger correlation with behaviour than intention.

  • Instrumental attitude had the strongest correlation with intention.

Abstract

This study investigates to what extent psychosocial factors, as identified by the theory of planned behaviour, determine the intention of the data producers to enter data correctly and to what extent their intention explains the actual errors in the entered data. The research goal is achieved by means of an observational study in a major Belgian financial institution. The results show that the degree of intention of the data producers to enter data correctly explains a certain share of the actual errors in the studied data, but that a larger share could be explained by statistical control variables related to actual behavioural control. In addition, the results indicate that the surveyed data producers’ intention towards entering high quality data is mainly determined by two factors: their instrumental and experiential attitude. This paper adds to the literature by (1) empirically evaluating the theory of planned behaviour in a context of manual data entry using not only self-reported measures but also a measure of actual behaviour, and (2) by demonstrating the importance of actual behavioural control.

Introduction

Many organisations are, or should be concerned with the quality of the data in their corporate databases. It is important that data in such corporate databases are of high quality not only to support regular business operations, but also to gain a competitive advantage [1], or to ensure regulatory compliance [e.g. with 2].

A large share of the data that resides in corporate databases are acquired manually [3, p. 11] most likely because automatisation of certain manual data entry procedures would be extremely difficult or next to impossible. For example, financial institutions have a hard time automating their manual procedures and thus often rely on manual data entry because of, for instance, their complex hierarchy of financial products which resulted in very complex business processes [4, p. 2]. Therefore, even when the ultimate goal should be to automate manual data entry processes, it is often not cost effective or feasible to do so, and certainly not in a short time span.

Many of the data that has been manually acquired contains errors (see e.g. [[5], [6], [7], [8], [9], [10], [11]]) which affect several dimensions of the data’s quality such as accuracy, correctness and consistency [12], [13]. Errors made when entering data are not always innocent: even a single one has the potential to cause serious issues. For example, an employee of a financial institution entered data erroneously which enabled a client to spend 2.1 million dollars before being caught [14, par. 55]. In another case, a prison guard made a keying error, causing the incorrect release of an inmate who was sentenced for stabbing someone [15]. In this same manner, typos made by US government employees cause about 500 people a month to be marked deceased when they are, in fact, alive [16].

To improve the quality of their manually acquired data, organisations typically undertake a combination of two types of initiatives. One type, data cleaning initiatives, aims to eliminate deficiencies that currently reside in a database by first identifying erroneous data, next choosing transformations to fix these data and finally applying the chosen transformations on the database [17]. However, if an organisation only cleans the data, it is still possible that new erroneous data are entered into the database [18, p. 564]. As such, the other type of initiatives that organisations undertake, preferably before cleaning the database [19, p. 55], are those that aim to eliminate the root causes of erroneous data input.

One of the root causes of errors in manually acquired data is that certain people who enter the data (i.e. data producers [20]) are not fully motivated to do this correctly [5]. In other words, the intention of some data producers to enter data correctly could be lacking.

To increase the intention of data producers to correctly enter data, organisations could initiate one or several behavioural interventions. In order for these interventions to be as effective as possible, they should target the appropriate psychosocial factors of the data producers [21].

The theory of planned behaviour [22] could provide an explanation of which psychosocial factors determine the intention of the data producers and eventually their actual behaviour of correctly entering the data [23]. This theory is well-established [24] and has been applied and empirically tested in a variety of different settings [25]. For example, [26] used the theory to investigate the behaviour of consuming fruit and vegetables, [21] employed it to change the behaviour of people into using public transport and [27] used it to study why people evade taxes. In an information systems setting, the theory is, for example, often used to explain the usage or the adoption of an information system artefact (IS artefact).

However, despite the plenitude of empirical tests in a variety of different settings, we see two reasons why additional evidence of the theory is desirable. First, in general, the link between intention and behaviour is not as often investigated as the link between the antecedents of intention, and even when this link is investigated, behaviour is almost always measured by self-reports [24, p. 486]. Second, in the context of manual data entry, the theory of planned behaviour has only been used for a theoretical argumentation on how data producers could be motivated to enter data of high quality [23]. Empirical evidence of the theory of planned behaviour in the context of manual data entry is lacking. The lack of empirical evidence also implies that it is not known which psychosocial factors are the most effective in strengthening the intention of data producers to enter data correctly.

In response, this study aims to answer the following research questions (RQs):

RQ 1

To what extent is the (actual) behaviour of data producers to enter data without errors determined by their intention to enter data without errors?

RQ 2

To what extent do the psychosocial factors of the theory of planned behaviour determine the data producers’ intention to enter data without error?

To answer these research questions, we empirically evaluate the theory of planned behaviour in a manual data entry context by means of a quantitative observational study in a major Belgian financial institution.

The remainder of this manuscript is structured as follows. Section 2 explains the basic concepts of the theory of planned behaviour. Section 3 presents a review of the literature which consists of (1) a benchmark of a representative set of studies in the information systems literature that empirically applied the theory of planned behaviour in contexts other than manual data entry, and (2) a discussion of how the benchmark results connect to other related research. Section 4 develops the hypotheses that will be tested in the empirical study. Section 5 delineates the methodology of the study. The results, discussion and conclusion will be presented in 6 Results, 7 Discussion and 8 respectively.

Section snippets

Theory of planned behaviour

The theory of planned behaviour (see Fig. 1) is a refinement of the theory of reasoned action [28] and states that the behaviour of individuals is determined by their intention and actual behavioural control (i.e. available resources and opportunities) to perform the behaviour [22]. The intention to perform the behaviour is, in its turn, explained by the individual’s attitude (beliefs about the likely consequences of performing the behaviour), subjective norm (beliefs about how significant

A quantitative benchmark of related studies

The theory of planned behaviour, and its predecessor are often applied and studied in the context of information systems, but, as explained above, not in the context of manual data entry. In order to be able to compare the results of our study to related research, we collected a set of information system studies that answered similar research questions and also used a quantitative research method. Concerning RQ 1, which is about the interaction between intention and behaviour, it would be

Entering data correctly: behavioural beliefs and hypotheses development

In this section we identify behavioural beliefs specific to manual data entry and develop hypotheses to investigate research questions RQ 1 and RQ 2. The behavioural beliefs and hypotheses are based on related literature and our experience during a preliminary exploratory study [i.e. 5].

Context

The hypotheses were tested in the context of the manual acquisition of home loan data in a major Belgian financial institution.

One of the most important attributes of a home loan is the value of the collateral (i.e. real-estate) that is used for the mortgage of the loan. This attribute is mainly used for determining the application credit score of the customer, the risk of the home loan and the regulatory capital required to cover this risk. The attribute is entered manually by home loan

Results

Table 4 shows the items of the questionnaire, the names of the variables and the descriptive statistics. The binary variable IsCorrect is used as the dependent variable in the logistic regression model. This variable is 1 if the value of the collateral for home loan of the chosen type was entered correctly and 0 otherwise. We found that the value of the collateral for 183 out of 204 home loans was entered completely correctly.

Discussion

The results of the logistic regression show that the intention of data producers to enter correct data is, as expected, positively correlated to the act of entering data correctly. This means that the amount of errors in this data is likely to be decreased when the data producers undergo a behavioural intervention aimed at increasing their intention to enter data correctly.

Yet, as indicated by the marginal effects analysis, the actual behavioural control of the individuals was probably more

Conclusion

In this paper we investigated to what extent the data producers’ intention to enter data without error explains the degree to which they actually enter data without error and to what extent their intention is determined by the psychosocial factors of the theory of planned behaviour. In our study, the degree to which the data producers intended to enter data without errors was able to explain a small share of the actual errors in the studied data. A larger share of errors was explained by

Acknowledgement

The authors would like to thank KBC Group NV for their financial and operational support.

References (103)

  • J. Jeffrey et al.

    Boosting safety behaviour: descriptive norms encourage child booster seat usage amongst low involvement parents

    Accid Anal Prev

    (2016)
  • I. Ajzen et al.

    Prediction of leisure participation from behavioral, normative, and control beliefs: an application of the theory of planned behavior

    Leis Sci

    (1991)
  • G.J. Fogarty et al.

    Safety climate and the theory of planned behavior: towards the prediction of unsafe behavior

    Accid Anal Prev

    (2010)
  • S. Goode et al.

    Exploring software piracy as a factor of video game console adoption

    Behav Inf Technol

    (2012)
  • S.-Y. Hung et al.

    User acceptance of mobile E-government services: an empirical study

    Gov Inf Q

    (2013)
  • Y. Hwang

    The moderating effects of gender on E-commerce systems adoption factors: an empirical investigation

    Comput Human Behav

    (2010)
  • F.-Y. Kuo et al.

    Predicting knowledge sharing practices through intention: a test of competing models

    Comput Hum Behav

    (2008)
  • T.-P. Liang et al.

    Contextual factors and continuance intention of mobile services

    Int J Mobile Commun

    (2013)
  • J. Lin et al.

    Understanding competing application usage with the theory of planned behavior

    J Am Soc Inf Sci Technol

    (2006)
  • M.-H. Huang et al.

    First-line and middle manager IT usage intention: a comparison of TAM, TRA and TPB

    Proceedings of the international conference on management and service science

    (2009)
  • M. Mäntymäki et al.

    Does a contextualized theory of planned behavior explain why teenagers stay in virtual worlds?

    Int J Inf Manage

    (2014)
  • M. Sojer et al.

    Understanding the drivers of unethical programming behavior: the inappropriate reuse of internet-accessible code

    J Manage Inf Syst

    (2014)
  • H.-E. Yang et al.

    The modified theory of planned behavior applied to E-government service acceptance

    Int J Serv Sci 2

    (2008)
  • T.C. Redman

    Improve data quality for competitive advantage

    Sloan Manage Rev

    (1995)
  • BCBS

    Principles for effective risk data aggregation and risk reporting

    Tech. Rep.

    (2013)
  • A. Maydanchik

    Data quality assessment

    (2007)
  • J. Dias et al.

    Automating the bank’s back office

    Tech. Rep.

    (2012)
  • Haegemans T, Snoeck M, Lemahieu W, Stumpe F, Goderis A. Towards a theoretical framework to explain root causes of...
  • N.G. Weiskopf et al.

    Methods and dimensions of electronic health record data quality assessment: enabling reuse for clinical research

    J Am Med Inform Assoc

    (2013)
  • N. DeHoratius et al.

    Inventory record inaccuracy: an empirical analysis

    Manage Sci

    (2008)
  • K. Thiru et al.

    Systematic review of scope and quality of electronic patient record data in primary care

    BMJ

    (2003)
  • D.G.T. Arts et al.

    Defining and improving data quality in medical registries: a literature review, case study, and generic framework

    J Am Med Inform Assoc

    (2002)
  • D.R. Goldhill et al.

    APACHE II, data accuracy and outcome prediction

    Anaesthesia

    (1998)
  • R.Y. Wang et al.

    Beyond accuracy: what data quality means to data consumers

    J Manage Inf Syst

    (1996)
  • R.W. Zmud

    An empirical investigation of the dimensionality of the concept of information

    Decis Sci

    (1978)
  • Moore v. R

    NSWCCA 260

    (2016)
  • Francis S. Man accused of stabbing clerk was accidentally released from jail early. 2017....
  • Bichell RE. Social Security Data Errors Can Turn People Into The Living Death. 2016....
  • V. Raman et al.

    Potter’s wheel: an interactive data cleaning system

    Proceedings of the international conference on very large databases

    (2001)
  • Y.U. Huh et al.

    Data quality

    Inf Softw Technol

    (1990)
  • T.C. Redman

    Data quality: the field guide

    (2001)
  • S. Bamberg

    Is a residential relocation a good opportunity to change people’s travel behavior? results from a theory-driven intervention study

    Environ Behav

    (2006)
  • I. Ajzen

    The theory of planned behavior

    Org Behav Hum Decis Process

    (1991)
  • G.D. Murphy

    Improving the quality of manually acquired data: applying the theory of planned behaviour to data quality

    Reliab Eng Syst Saf

    (2009)
  • J. Brug et al.

    Psychosocial determinants of fruit and vegetable consumption.

    Appetite

    (1995)
  • D.J. Hessing et al.

    Exploring the limits of self-reports and reasoned action: an investigation of the psychology of tax evasion

    J Pers Soc Psychol

    (1988)
  • M. Fishbein et al.

    Predicting and changing behavior: the reasoned action approach

    (2010)
  • F.D. Davis

    A technology acceptance model for empirically testing new end-user information systems

    (1985)
  • F.D. Davis

    Perceived usefulness, perceived ease of use, and user acceptance of information technology

    MIS Q

    (1989)
  • V. Venkatesh et al.

    User acceptance of information technology: toward a unified view

    MIS Q

    (2003)
  • Cited by (4)

    • A theoretical framework to improve the quality of manually acquired data

      2019, Information and Management
      Citation Excerpt :

      However, the importance of the TTF or actual behavioural constructs, which are conceptually equal in the context of manual data entry, is often neglected when behavioural theories are applied in an information system context. For example, in a recent overview of studies that empirically and quantitatively tested the theory of planned behaviour in an information system context, none of the studies included measures for actual behavioural control [27]. In addition, several frequently cited individual information system theories such as the technology acceptance model [55], do not include constructs that capture the degree to which individuals have control over using information system artefacts [24].

    • University Teacher Service Platform Integrated with Academic Social Network

      2021, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    • Data Quality Management: An Overview of Methods and Challenges

      2021, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    View full text