Search

Statistical methods play an important role in auditors’ analyses of auditee’s data. A key aspect in the statistical approach to auditing is assessing the strength of evidence for or against a hypothesis. Unfortunately, the often-used frequentist statistical methods cannot provide the statistical evidence that audit standards demand directly nor easily. In this article we discuss an alternative approach that can provide this evidence: Bayesian inference. Firstly, we explore the philosophical differences between frequentist and Bayesian inference. Secondly, we discuss misconceptions in the interpretation of frequentist statistical evidence, and finally we discuss how Bayesian inference allows the auditor to obtain and interpret statistical evidence in line with audit standards via its alternative to the p value, the Bayes factor.

The audit environment of today offers a wealth of information in the form of data. Consequently, data about the auditee is expected to guide and improve auditors’ approach to tests of details. However, to be able to make optimal use of this data, auditors must have tools that facilitate the effective and efficient use of quantitative information throughout an audit. In this article, we introduce Bayesian generalized linear modeling as a statistical framework to incorporate this information into tests of details, thereby enabling auditors to deliver a fine-grained and specifically tailored audit opinion to stakeholders. We begin with an introduction of Bayesian inference in audit sampling, then explain the main concepts underpinning Bayesian generalized linear modeling and show how this approach allows auditors to bridge the gap between analytics on integrally available data and analytics on data that is available on a sample basis, making optimal use of their information.

We outline a network method to synthesize a literature from search results obtained by multiple team members. Several network statistics are used to create a single representativeness ranking. We illustrate the method with the dispersed literature on a common misinterpretation of analysis of covariance (ANCOVA). The network method yields a top ten list of most relevant articles that students and researchers can take as a point of departure for a more detailed study on this topic.

This project will outline the development and application of a new Bayesian assumption check. The purpose of this check is to assess deviations from normality of the distribution of a variable. Adding this check to the Bayesian arsenal of tests is important because it will let researchers know if their data deviates from normality and thus need to consider Bayesian non-parametric methods. During this process I will define a Bayes factor that quantifies deviations from normality, by comparing a model that assumes normality to a model that incorporates specific deviations from normality. Several candidate models fit the latter description and are therefore part of the main interests of this research. A comparison between the candidate models will be given and an assessment of their appropriateness for the test will be made. If there is have time remaining, I will apply the developed test in a linear regression setting.

Bayesiaanse statistiek is een manier om efficiëntie en transparantie bij gestratificeerde steekproeven te vergroten. De reden hiervoor is dat Bayesiaanse statistiek de auditor dwingt om expliciet te maken welke informatie en kennis gebruikt wordt bij de evaluatie van de steekproef. In dit artikel worden twee belangrijke vormen van voorkennis beschreven die de auditor moet valideren. Vervolgens zijn deze vormen van voorkennis vertaald naar statistische modellen, die worden gebruikt bij het doen van een gestratificeerde steekproef. Afsluitend wordt gedemonstreerd hoe sommige van die modellen kunnen leiden tot nauwkeurigere foutschattingen en een transparantere audit.

Auditors who perform audit sampling are often interested in obtaining evidence for or against the hypothesis that the misstatement in a population of items is lower than a critical limit, the so-called performance materiality. Here, we propose to perform this hypothesis test using a Bayesian approach that involves the use of an impartial prior distribution, assigning equal prior probabilities to the competing interval hypotheses. Firstly, we argue that the impartial prior distribution is sensible for auditors because it is easy to justify, interpret, and explain. Secondly, we show that Bayes factors computed using this prior distribution have desirable statistical properties. Finally, we compare these Bayes factors with traditional p-values in an audit sampling context and elaborate on the merits of the impartial Bayesian hypothesis test.

Analysis of variance (ANOVA) is the standard procedure for statistical inference in factorial designs. Typically, ANOVAs are executed using frequentist statistics, where p-values determine statistical significance in an all-or-none fashion. In recent years, the Bayesian approach to statistics is increasingly viewed as a legitimate alternative to the p-value. However, the broad adoption of Bayesian statistics-and Bayesian ANOVA in particular-is frustrated by the fact that Bayesian concepts are rarely taught in applied statistics courses. Consequently, practitioners may be unsure how to conduct a Bayesian ANOVA and interpret the results. Here we provide a guide for executing and interpreting a Bayesian ANOVA with JASP, an open-source statistical software program with a graphical user interface. We explain the key concepts of the Bayesian ANOVA using two empirical examples.

Adolescence compared to childhood and adulthood, is characterized by more intense positive and negative emotions (Verma, & Larson, 1999; Silk, Steinberg, & Morris, 2003), more dynamic emotional states (Somerville, Jones, & Casey, 2010), and greater sensitivity to social and emotional stimuli (Spear, 2009; Bailen, Green, & Thompsom, 2019). In 2020, the COVID-19 pandemic radically changed the already emotionally rich lives of adolescents worldwide, presenting them with unprecedented challenges for their emotional wellbeing. To navigate the landscape of fluctuating emotions and to adapt to daily life under the extraordinary circumstances introduced by the pandemic, adolescents need emotion regulation. At present, few studies have investigated emotion regulation processes in daily life during pandemics or how these processes are related to psychopathology and contextual factors such as intensity of daily positive and negative events. In the current study, we will investigate rumination, savoring, and sharing in the daily life of adolescents during the social distancing measures put in place in Belgium due to the COVID-19 pandemic. Specifically, we will assess if current psychopathology and event intensity are associated with rumination and sharing in response to negative daily events, and with savoring and sharing in response to daily positive events. To do so, we will use data from an extra measurement within the ongoing SIGMA study (Kirtley et al., 2021; Achterhof et al., 2021), conducted with 173 adolescents from the SIGMA cohort in Flanders (Belgium). Data were collected between 04.05.2020 and 10.05.2020, during the first COVID-19 related lockdown in Belgium. In this extra measure within the SIGMA study, psychopathology was assessed via a self-report questionnaire as a part of a large battery of self-report questionnaires, filled in online at home. Intensity of daily positive and negative events, rumination, sharing, and savoring were measured during a six-day ESM (Experience Sampling Method) period as a part of a momentary ESM survey delivered to the participants 10 times per day. As we have also recently conducted a daily life study with data collected before the pandemic where the same variables were assessed in daily life with similar methodology and the same emotion regulation items, we consider the current study a conceptual replication (i.e., re-testing the same hypotheses again with a different sample, study design, time-period, or measurement method) of previous research (Hiekkaranta et al., 2021). Replication studies are rare in the field of emotion regulation and we are not currently aware of any daily life replication study in this area. Replication studies, though rarely conducted, are vital in testing both the robustness and universality of effects observed in previous studies. Therefore, given the nature of the current conceptual replication study, we derived hypotheses from the results of the study that we aim to replicate.