Nudging auditors' unconscious to improve performance on an accounting estimate task

• Published date: 01 April 2022
• Author: Christine J. Nolder,Nicole V. S. Ratzinger‐Sakel,Jochen C. Theis
• Date: 01 April 2022
• DOI: http://doi.org/10.1111/ijau.12252

ORIGINAL ARTICLE

Nudging auditors' unconscious to improve performance on an

accounting estimate task

Christine J. Nolder

1

| Nicole V. S. Ratzinger-Sakel

2

| Jochen C. Theis

3

1

Accounting Department, Suffolk University,

Boston, Massachusetts, USA

2

Chair of Accounting and Auditing, University

of Hamburg, Hamburg, Germany

3

Department of Business and Management,

University of Southern Denmark, Kolding,

Denmark

Correspondence

Jochen C. Theis, Department of Business and

Management, University of Southern

Denmark, Kolding, Denmark.

Email: jot@sam.sdu.dk

Auditing deficiencies related to accounting estimates continue to account for a signif-

icant percentage of auditing deficiencies identified during routine inspections. Regu-

lators suggest that a lack of auditors' professional skepticism may be to blame. We

test a skeptical language priming intervention to induce auditors into a skeptical

mindset. Additionally, we posit that skeptical language embedded in a firm's decision

aid will induce a skeptical attitude, which we measure in terms of auditors' beliefs

and feelings (i.e., confidence and concern) associated with the reasonableness of

management's estimate. We find that the skeptical language prime has a positive

effect on the quality of auditors' cognitive processing and reasonableness judgments.

We also find that the skeptical language keeps auditors' feelings of concern high as

opposed to conventional language, which resulted in auditors feeling less concerned.

Moreover, we found that auditors' negative feelings of concern were the primary

driver of recommended testing both before and after introducing the firm's

decision aid.

KEYWORDS

accounting estimates, mindset and attitude, professional skepticism

1|INTRODUCTION

Despite a growing trend of improved audit quality worldwide,

auditing deficiencies related to accounting estimates continue to

comprise a significant percentage of auditing deficiencies identified

during routine inspections (International Forum of Independent Audit

Regulators (IFIAR), 2017, 2018, 2019; Public Company Accounting

Oversight Board (PCAOB), 2018a). The auditing deficiencies repre-

sent situations where auditors did not perform enough testing to

comply with the relevant auditing standard. For example, PCAOB-

identified auditing deficiencies associated with AS2501 Auditing

Accounting Estimates often describe situations where firms' proce-

dures were limited to inquiry (Public Company Accounting Oversight

Board (PCAOB), 2019a, 2019b, 2019c, 2019d). Regulators investi-

gating potential root causes of these auditing deficiencies suggest

that a lack of auditors' professional skepticism may be to blame

(Public Company Accounting Oversight Board (PCAOB), 2018a;

International Forum of Independent Audit Regulators (IFIAR), 2015,

2016, 2017).

If auditors' professional skepticism (or lack thereof) is, in fact, a

primary driver of accounting estimate-related auditing deficiencies,

firms are required to design remediation strategies aimed at improving

the likelihood that auditors plan and perform audits with the requisite

levels of professional skepticism (Public Company Accounting

Oversight Board (PCAOB), 2018a; International Auditing and

Assurance Standards Board (IAASB), 2020). To illustrate, the Interna-

tional Standard on Quality Management (ISQM 1) requires that firms

‘respond to circumstances when findings indicate that there is an

engagement(s) for which procedures required were omitted during

the performance of the engagement(s)’(International Auditing and

Assurance Standards Board (IAASB), 2020, para. 45). Moreover, firms

must document remedial actions taken to address the identified

auditing deficiencies and include evidence (e.g., published auditing

research) to support the effectiveness of such actions (International

Auditing and Assurance Standards Board (IAASB), 2020). Thus,

understanding how to effectively manage requisite levels of auditors'

professional skepticism to improve audit quality is a relevant and

important topic in practice that warrants researchers' attention.

Received: 10 February 2021 Revised: 13 October 2021 Accepted: 13 October 2021

DOI: 10.1111/ijau.12252

78 © 2021 John Wiley & Sons Ltd Int J Audit. 2022;26:78–93.wileyonlinelibrary.com/journal/ijau

From an academic perspective, designing an effective remediation

strategy necessitates defining professional skepticism and identifying

theories that explain how it can be measured, monitored, and

improved. Consistent with Nolder and Kadous (2018), we view pro-

fessional skepticism as comprising two components, a skeptical

mindset and a skeptical attitude. The mindset component is reflected

in auditors' cognitive processing measures (e.g., mental representa-

tion, information search), while the attitude component is reflected in

auditors' cognitive (e.g., beliefs about reasonableness) and affective

(e.g., feelings of confidence and concern) evaluative judgments

associated with financial statements (e.g., risk judgments) and what

constitutes appropriate sufficient evidence. Since attitudes are

often theorized as the proximal antecedent to intentions (Ajzen &

Fishbein, 1980; Fishbein & Ajzen, 1975), understanding auditors'

skeptical attitudes will increase our ability to predict their intentions

to perform additional testing necessary to comply with the standards.

Using both mindset and attitude conceptualizations of profes-

sional skepticism, the objective of this research is two-fold. First,

we test a priming intervention recommended by Nolder and

Kadous (2018) to induce auditors into a skeptical mindset. More spe-

cifically, we embed skeptical language in a firm's decision aid (e.g., an

audit firm's guidance on how to perform certain audit procedures) that

is theorized to induce a skeptical mindset and positively influence

auditors' skeptical attitudes. Second, we explore whether auditors'

cognitive evaluative judgments (i.e., beliefs about reasonableness) or

affective evaluative judgement (i.e., feelings of confidence and con-

cern) are more indicative of auditors' intentions to perform additional

testing, which will provide insights about the likely effectiveness of

remediation designs aimed at auditors' cognitive versus affective

responses to risks.

To accomplish our objectives, we conducted an experiment with

58 experienced auditors from three Big4 audit firms.

1

Our audit case

includes financial information associated with a significant accounting

estimate for a fictitious German electronics manufacturer. After read-

ing the case, we ask participants a number of questions to capture the

nature and extent of their cognitive processing (i.e., measures of a

skeptical mindset), their cognitive and affective evaluative judgments

(i.e., measures of a skeptical attitude), and their recommendation for

additional testing (i.e., measure of intention). Collecting these mea-

sures before (i.e., at Time 1) introducing a priming intervention (i.e., at

Time 2) allows us to establish a baseline for auditors' initial gut-

feelings that drive their judgments and decisions (see Humphrey &

Moizer, 1990; Olsen, 2017). Then, at Time 2, we can measure the

extent to which firms' decision aids affect auditors' cognitive

processing and shift the salience of auditors' affective versus cognitive

evaluative judgments.

When we introduced the priming intervention, half of the

participants received a firm's decision aid with embedded skeptical

language based on Nolder and Kadous's (2018) recommendation

(i.e., ‘receptive’,‘open’,‘alert’,‘objective’and ‘unbiased’).

2

The other

half of the participants received a firm's decision aid with embedded

conventional language in ISA 540, hereafter referred to as baseline

language (i.e., ‘test’,‘review’and ‘determine’; International Auditing

and Assurance Standards Board (IAASB), 2018).

3

After reading the

firm's decision aid and completing the recommended procedures, we

asked auditor participants the same set of questions measuring their

skeptical mindset, skeptical attitude, and intentions to perform addi-

tional procedures to test the reasonableness of management's esti-

mate. We predict auditors in the skeptical language condition to

exhibit a higher level of professional skepticism in terms of both

mindset (H1a, b) and attitude (H2a, b) measures compared to those in

the baseline language condition.

We find that participants in the skeptical language condition

exhibited higher quality cognitive processing compared to baseline

language participants as measured by a higher number of valid addi-

tional supporting and non-supporting (i.e., challenging) arguments

regarding management's estimate after reading the firm's decision aid

(H1b). Participants in the skeptical language condition also exhibited

higher quality evaluative judgments as measured by their assessed

reasonableness of management's estimate (cognitive evaluative judg-

ments; H2a) and expressed concern about the risk of a material

misstatement (affective evaluative judgments; H2b). Moreover, the

participants in the skeptical language condition exhibited greater con-

sensus (i.e., less variation) in their reasonableness judgments, which

some audit researchers argue is a proxy for higher judgement quality

(see Ashton, 1974; Bonner, 2008).

We next explore whether the recommended testing at Time

1 and Time 2 can be explained by auditors' beliefs about reasonable-

ness (i.e., cognitive evaluative judgement) or positive/negative feelings

of confidence/concern (i.e., affective evaluative judgement), respec-

tively (RQ1). Because firms' decision aids, in general, are known for

being a cognition-focused intervention strategy (Larrick, 2004), we

expect that introducing a firm's decision aid to participants will

increase the correspondence between auditors' cognitive evaluative-

judgments and their recommended testing. However, we do not

include a formal hypothesis since we are unsure of the effect of a

firm's decision aid on auditors' affective evaluative-judgments. We

find that at Time 1, auditors' affective evaluative judgments

(i.e., feelings of confidence and concern) are significantly associated

with their judgments about recommended testing, while their beliefs

about reasonableness are not. These results are consistent with the

findings in the risk communication research that suggest emotional

responses are more predictive of behaviour (actual or intended) in a

risk setting (Loewenstein et al., 2001; Slovic et al., 2002, 2004). How-

ever, introducing the firm's decision aid decreased the influence of

auditors' confidence and increased the influence of auditors' cognitive

evaluative judgments (i.e., beliefs about reasonableness) on auditors'

recommended testing at Time 2 (with no significant difference

between the skeptical and baseline language condition). While there

are a growing number of papers that advocate for unconsciously

influencing auditors' judgments through priming interventions

(e.g., Bucaro, 2019; Griffith, Hammersley, Kadous, & Young, 2015;

Rasso, 2015), our study suggests that while priming auditors' skeptical

mindset or skeptical attitude may improve their reasonableness judg-

ments, it may not drive the recommended testing, as we do not find a

significant difference between conditions. Future intervention studies

NOLDER ET AL.79