The global introduction of Audit Quality Indicators (AQIs) aims to improve audit quality by offering new insights that enhance transparency and comparability between audits and audit firms (FEE 2016; Chen et al. 2018; Pinello et al. 2019). These frameworks provide visibility into the audit process, which can directly influence investors’ perception of a company’s financial statement reliability (Brown and Popova 2019; Chen et al. 2018; Foley 2025; Pinello et al. 2019).

To provide guidance, the International Auditing and Assurance Board (IAASB) framework breaks down Audit Quality (AQ) into five groups: inputs, process, outputs, interactions, and context (Christensen et al. 2015; FEE 2016). The inputs category covers crucial elements like organizational culture, auditors’ knowledge, skillset, and time spent on the engagement (FEE 2016).

The fundamental challenge in audit disclosures is the expectation gap, which describes the divergence between the public’s expectations of auditor responsibilities and the auditor’s own definition of their role (Christensen et al. 2015; Church et al. 2008; Koh and Woo 1998; Ruhnke and Schmidt 2014). Audit professionals define quality primarily as compliance with professional auditing standards, focusing on the outputs and process categories (Christensen et al. 2015; FEE 2016). In sharp contrast, investors rank the inputs category as the most valuable information, prioritizing AQIs related to the composition, quality, and experience of the audit team (Christensen et al. 2015; FEE 2016).

Publishing engagement team AQI disclosures, particularly those focused on inputs, holds the potential to reduce this gap by increasing transparency on and bridging information asymmetry (Brown and Popova 2019; The Center for Audit Quality 2014). Current misunderstandings often arise through investors’ difficulty in assessing auditor performance, resulting in investors holding auditors to a higher standard of care (Ruhnke and Schmidt 2014). However, audit organizations emphasize the necessity of accurate and comprehensive information, warning that superficial frameworks could lead to misinterpretation (Brown and Popova 2019; Foley 2025; Koops-Aukes 2021).

Academic literature supports the informational value of Audit Quality Indicator (AQI) disclosures, particularly when they provide insight into audit inputs that are otherwise unobservable to financial statement users. Prior research shows that audit quality signals can reduce both type I (false positives) and type II (false negatives) investment errors by lowering information asymmetry and improving investors’ ability to assess financial reporting reliability (Chen et al. 2018; Christensen et al. 2015). Input-based AQIs, such as auditor expertise, training, and engagement team involvement, are especially valued by investors, who associate these attributes with higher audit quality and greater reporting credibility (Christensen et al. 2015; FEE 2016).

However, the effect of AQI disclosures on investor perception is not uniformly positive. Evidence from related assurance contexts indicates that investor responses depend on the content and context of the disclosed information. For example, research on Internal Audit Reports demonstrates that disclosures regarding assurance activities and committee responsibilities can either enhance or diminish investor trust, depending on how the information is interpreted (Holt and DeZoort 2009). This suggests that transparency does not inherently increase confidence; rather, the nature of the disclosed audit inputs determines their informational value.

Building on this insight, the present study focuses on two engagement-level, input-based AQIs: the level of partner and manager involvement and the disclosure of auditor training hours. Higher involvement by senior audit personnel is generally interpreted as a positive signal of audit effort and oversight, whereas lower involvement may raise concerns about audit rigor. At the same time, training-related disclosures may provide complementary information about the competence and expertise of the engagement team, potentially contextualizing or mitigating concerns arising from lower involvement levels of the more senior team members. Accordingly, the effects of AQI disclosures are expected to depend on how multiple audit inputs jointly shape investor judgments, rather than on the mere presence of disclosure.

Based on this reasoning, the study develops hypotheses that focus on the relative effects and interaction of specific AQI disclosures, rather than predicting a uniform effect of AQI disclosure compared to the absence of such information.

Within the input category partner and manager involvement is an observable factor heavily relied upon by non-professional investors when assessing audit quality (Christensen et al. 2015; Church et al. 2008). This involvement operates through signaling theory, where senior-level participation sends a credible signal of quality (Harris and Williams 2020; Spence 1978). Empirical evidence confirms that investors place significant value on the involvement of senior audit professionals, associating it with higher audit quality (Christensen et al. 2015). This leads to the following hypotheses:

H1a: Disclosures indicating higher partner and manager involvement, compared to lower partner and manager involvement, will lead investors to perceive higher financial statement reliability.

H1b: Disclosures indicating higher partner and manager involvement, compared to lower partner and manager involvement, will lead to higher investment likelihood.

A major concern voiced by audit firms is that heavily input-based disclosures are only meaningful when they provide a more detailed view of team characteristics, rather than being isolated indicators (Koops-Aukes 2021; KPMG Accountants NV 2021; PwC et al. 2021). For instance, audit firms fear investors may place too much weight on partner and manager involvement when disclosures lack context (Foley 2025). This has led to specific professional calls, such as those by Koops-Aukes (2021) to the AFM in the Netherlands, to incorporate a training hours disclosure alongside existing mandatory metrics to provide a more comprehensive view of engagement team quality (Koops-Aukes 2021; KPMG Accountants NV 2021; PwC et al. 2021).

Academically, these concerns are supported by the theory that perceived audit-quality judgments depend on whether investors see a comprehensive framework of signals (Chen et al. 2018). Furthermore, additional explanations provide greater informational value, particularly when the initial AQI provides an ambiguous or negative signal (Baginski et al. 2004; Hutton et al. 2003; Mercer 2004; Skinner 1994). To test the effect of the addition of a training related disclosure as a potential remedy to mitigate potential ambiguous or negative signals, the following hypotheses were tested:

H2a: The effect of a training hours disclosure on perceived financial statement reliability depends on the level of partner and manager involvement, such that the disclosure increases perceived reliability when partner and manager involvement is low, but has no effect when partner and manager involvement is high.

H2b: The effect of a training hours disclosure on investment likelihood depends on the level of partner and manager involvement, such that the disclosure increases investment likelihood when partner and manager involvement is low, but has no effect when partner and manager involvement is high.

Training-related disclosures are conceptually suitable additions because they signal competence and a firm’s investment in human capital (Christensen et al. 2015). Training reflects staff competence and is one of the most frequently adopted AQIs globally, making it particularly useful when another input-based AQI, like the lack of partner and manager involvement, might instill ambiguity in investors’ perception of audit quality (FEE 2016; Harris and Williams 2020; Koops-Aukes 2021; Spence 1978).

The core theory is that this contextual information is most effective when the primary signal is weaker (Baginski et al. 2004; Brown and Popova 2019). Thus, when partner and manager involvement is low, a training hours disclosure is expected to mitigate non-professional investors’ perception of low audit quality. Conversely, if involvement is already high, the added training disclosure may be redundant since the positive quality signal is already established (Baginski et al. 2004; Hutton et al. 2003; Skinner 1994). Incorporating a training related disclosure aligns the Dutch AQI framework with international practices and addresses firm concerns regarding comprehensiveness.

This study utilized a 2 (lower vs. higher partner and manager involvement) × 2 (training hours present vs. absent) + 1 (control condition: no information about partner and manager involvement or training hours) between-subjects design to examine how investors react to specific audit quality disclosures.¹

The sample consisted of 108 business- and accounting-related students, proxying for reasonably informed retail investors, alongside a minority of non-professional investors. This use of students as proxies is supported by extensive prior literature, which suggests that business and accounting students can accurately replicate non-professional investor behavior in low-complexity, judgment-based experiments (Ashton and Kramer 1980; Elliott et al. 2007; Libby et al. 2002).

The sample size of 108 participants was sufficient for the setting, meeting the recommended minimal requirement of 20 participants per condition to ensure power and accuracy during analysis (Simmons et al. 2011; Van Voorhis and Morgan 2007). According to Table 1, the vast majority of respondents reported a business background (98%) and investment experience (74%).

The experiment was administered online, and participants were randomly assigned to one of the five conditions. Participants assumed the role of a non-professional investor evaluating a fictional company, GreenSip Group (Appendix 1). Before proceeding, comprehension check questions were administered to ensure participants understood the information, which is critical for ensuring the validity, accuracy, and reliability of self-reported data (Brener et al. 2003; Christensen et al. 2014).

Both the Independent Variable (IV) and the Moderating Variable (MV) disclosures were developed according to specific guidelines made by the PCAOB (Vanich et al. 2024).

Data analysis began after cleaning and excluding participants who failed manipulation checks.

A 2x2 factorial ANOVA was initially planned to test the hypotheses, which addressed the main and interaction effects of the variables (Statistics Solutions n.d.). However, this test relies on the fundamental assumption that the variances between the conditions are equal (homogeneity of variance). During the data analysis, Levene’s test revealed that significant heterogeneity of variances was found across conditions for hypotheses 1 and 2. Consequently, due to this violation of the homogeneity assumption, the ANOVA test was deemed less suitable for providing statistically valid results. To ensure statistical validity and to properly evaluate both the main and the interaction effects of the IV and MV on the dependent variables, the analysis was shifted to utilize individual t-tests and a linear regression with heteroskedasticity-robust standard errors (Dobson 2002). This strategy better aligns with best practices for this between-subjects design and provides a valid inference for testing simple and moderating effects across all conditions (Dobson 2002).

Manipulation checks confirmed the successful manipulation of both the Independent Variable (IV) and the Moderating Variable (MV). Participants in the high partner and manager involvement condition rated involvement significantly higher than the participants in the low involvement condition (p < 0.0001, not tabulated), confirming that participants correctly interpreted the manipulation. Furthermore, participants that were shown the moderating variable (training hours disclosure), were asked to rate 90 hours of training on a scale of -5 to 5. Participants that were shown the MV significantly interpretated the 90 hours of training as positive (M > 0, not tabulated), confirming their correct understanding of the disclosure (p < 0.001).

Random assignment was successful for the majority of participant characteristics. No significant differences were observed across experimental conditions for gender, investment experience, or business background. Differences did emerge for age and familiarity with audit reports, despite random assignment. These differences, therefore, reflect random imbalance rather than systematic selection. Importantly, all participants evaluated identical audit reports, with only the AQI disclosures manipulated, which limits the likelihood that these participant characteristics drive the observed effects. Nevertheless, these imbalances should be taken into account when interpreting the results and are acknowledged as a limitation of the study.

A Pearson correlation test between DV financial reporting reliability and DV investment likelihood showed a significant and moderately strong positive relationship between both dependent variables (r = 0.648, p < 0.001, not tabulated). This means that participants who assessed perceived financial reporting reliability as higher also assigned a higher score to investment likelihood. This indicates that investors who deem financial statements reliable are also more likely to invest in the company.

Hypotheses 1a and 1b predict that the level of partner and manager involvement (high versus low) influences both investors’ perceived financial reporting reliability and their likelihood of investing. To examine these hypotheses, independent sample t-tests were conducted.

H1a predicts that issuing a disclosure on partner and manager involvement (high or low) affects investors’ perception of financial reporting reliability. A t-test was used to examine the perception of financial reporting reliability between participants shown high and low involvement. Results showed that participants in the high involvement group significantly perceived financial reporting reliability as higher compared to participants in the low involvement group (M = 3.45 versus M = 0.44, p = 0.0004). As shown in Table 2 the result of the t-test confirms H1a. Issuing a disclosure showcasing high partner and manager involvement leads to a significantly higher perception of financial reporting reliability. Contrarily issuing disclosures showcasing low partner and manager involvement leads to significantly lower perception of financial reporting reliability.

Hypothesis 1b examined the effect of low and high partner and manager involvement on investment likelihood. A t-test was used to examine the effect on investment likelihood between participants shown high or low partner and manager involvement.

According to Table 3, participants shown high partner and manager involvement were significantly more likely to invest compared to participants shown low involvement (M = 2 versus M = -0.667, p = 0.0024).

The results confirm H1b, indicating that investors are more likely to invest when shown a disclosure indicating higher partner and manager involvement compared to a disclosure indicating lower partner and manager involvement.

The abovementioned results indicate that issuing a disclosure showcasing high partner and manager involvement increases both perceived financial reporting reliability and investment likelihood.

Hypotheses 2a and 2b examined whether the addition of a training hours disclosure influences investor behavior and whether these effects depend on the level of partner and manager involvement. To test these hypotheses a linear regression analysis, including interaction terms, was conducted.

The findings reveal that the use of AQI disclosures significantly enhances investors’ perception of financial reporting reliability. Disclosures showcasing high partner and manager involvement strongly enhanced perceived reliability and investment likelihood, confirming earlier findings (Chen et al. 2018; Christensen et al. 2015; Holt and DeZoort 2009). Conversely, low partner and manager involvement significantly lowered scores for the dependent variables, underscoring that investors view the role of senior auditors as a credible indicator of audit quality (Church et al. 2008; Foley 2025).

The overall addition of AQIs did not significantly increase perceived reliability; in fact, perceived negative AQI trends (i.e., low involvement) yielded a significantly lower perception of reliability compared to the control condition. This highlights that AQIs containing information perceived negatively can diminish trust when not accompanied by additional contextual information (Church et al. 2008).

Crucially, the addition of the training hours disclosure only had a significant positive effect on perceived financial reporting reliability when combined with low partner and manager involvement. This effect, however, was not significant for investment likelihood, though it approached significance. The moderate positive correlation between reliability and investment likelihood indicates a strong relationship between the two dependent variables, but the divergence in the training disclosure’s effect suggests that training information may be more useful as a signal for credibility rather than having a direct effect on investment behavior (Harris and Williams 2020; Spence 1978).

For policymakers and standard setters in the Netherlands, the results highlight the importance of providing comprehensive disclosure content. Since the Dutch AQI currently lacks a mandatory training hours disclosure, the results demonstrate the potential value of adopting this metric to signal quality and address investor concerns. The findings also justify audit firms’ argument that disclosures must be comprehensive to mitigate the negative effects of single, potentially adverse metrics (Foley 2025). Moreover, the research indicates that investors should critically assess all engagement disclosures collectively, rather than relying on any single disclosure.

Academically, this research contributes new empirical results on the effect of AQIs, expanding the existing literature by analyzing specific AQIs in a Dutch context (Brown and Popova 2019; Chen et al. 2018). Furthermore, the study validates claims derived from the expectations gap and signaling theory, demonstrating how specific audit inputs are interpreted and how these interpretations influence investor behavior (Christensen et al. 2015; Church et al. 2008; Harris and Williams 2020; Ruhnke and Schmidt 2014; Spence 1978).

This research acknowledges limitations, including the use of student proxies, which may limit generalizability, and the controlled experimental nature of the study, which limits external validity. Furthermore, a randomization check indicated that not all demographic variables were successfully randomized across groups.

A key methodological limitation was the analytical constraint imposed by unequal variances across condition groups (heteroskedasticity). This violation of the homogeneity assumption necessitated the use of linear regression with robust standard errors instead of the intended ANOVA test (Statistics Solutions, n.d; Dobson 2002).

Future research should expand the pool of AQIs tested (e.g., workload, tenure) and further investigate why the moderating effect of the training disclosure yielded different results across the two dependent variables. Broadening the regulatory environment and using actual investors as participants would further increase the generalizability and external validity of the results.

In conclusion, this article provides essential evidence on how AQIs influence investor behavior in the Netherlands, underscoring the value of providing comprehensive disclosures and aligning audit practices with investor expectations.