Explainability Versus Accuracy of Machine Learning Models - The Role of Task Uncertainty and Need for Interaction with the Machine Learning Model

Aug 5, 2026

This paper investigates the importance of explainability versus accuracy of machine learning (ML) models, focusing on how task uncertainty shapes interaction needs and shifts the relative value of explanations. It examines ML use for product cost estimation in new product development and applies mixed-method evidence to analyze the trade-off between accuracy and explainability. Findings from cost experts support an inverse relationship between explainability and accuracy. The paper also shows that although a complex and accurate gradient boosted regression model (GBR) is available, it is preferred only in a few situations; in most cases, more basic but better-explainable models are chosen despite lower accuracy. Additional analysis identifies factors that raise task uncertainty, increasing explainability importance: project unpredictability, product cost granularity, predecessor product availability, target cost gap, and development phase.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages35

Reading Time1 h 28 min

AccessFree to download

Introduction

Explainability-accuracy trade-off

Explainability needs in managerial accounting

Study context and approach

Product cost estimation in new product development

Mixed-methods evidence

Key findings

Inverse relationship from cost experts

Model preferences across situations

Task uncertainty drivers

Project unpredictability

Cost granularity and related factors

European Accounting Review ISSN: (Print) (Online) Journal [homepage: ](homepage: www.tandfonline.com/journals/rear20)[www.tandfonline.com/journals/rear20](homepage: www.tandfonline.com/journals/rear20) Explainability Versus Accuracy of Machine Learning Models: The Role of Task Uncertainty and Need for Interaction with the Machine Learning Model Dominik Hammann & Marc Wouters To cite this article: Dominik Hammann & Marc Wouters (25 Feb 2025): Explainability Versus Accuracy of Machine Learning Models: The Role of Task Uncertainty and Need for Interaction with the Machine Learning Model, European Accounting Review, DOI: 10. 1080/09638180 .2025.2463961 To link to this article: [https://doi.org/10.1080/09638180.2025.2463961](https://doi.org/10.1080/09638180.2025.2463961) © 2025 The Author(s) . Published by Informa UK Limited, trading as Taylor & Francis Group. | View supplementary material | | | --- | --- | | Published online: 25 Feb 2025. | | | Submit your article to this journal | | | | Article views: 382 | | | View related articles | | View Crossmark data | | Full Terms & Conditions of access and use can be found at [https://www.tandfonline.com/action/journalInformation?journalCode=rear20](https://www.tandfonline.com/action/journalInformation?journalCode=rear20) European Accounting Review, 2025 [https:](https://doi.org/10.1080/09638180.2025.2463961)[//](https://doi.org/10.1080/09638180.2025.2463961)[doi.org](https://doi.org/10.1080/09638180.2025.2463961)[/](https://doi.org/10.1080/09638180.2025.2463961)[10.1080](https://doi.org/10.1080/09638180.2025.2463961)[/](https://doi.org/10.1080/09638180.2025.2463961)[09638180.2025.2463961](https://doi.org/10.1080/09638180.2025.2463961) Explainability Versus Accuracy of Machine Learning Models: The Role of Task Uncertainty and Need for Interaction with the Machine Learning Model DOMINIK HAMMANN∗ and MARC WOUTERS ∗ , ∗∗ ∗ Department of Economics and Management, Karlsruhe Institute of Technology, Karlsruhe, Germany ∗∗ Amsterdam Business School, University of Amsterdam, Amsterdam, Netherlands (Received: April 2023; accepted: January 2025) Abstract This paper investigates the importance of explainability versus accuracy of machine learning (ML) models. We propose that greater task uncertainty makes people want to interact more with the ML model, which increases the importance of explainability relative to accuracy. We focus on the use of ML models for product cost estimation during new product development. The paper provides mixed-methods evidence on the trade-off between explainability and accuracy of ML models. Speciﬁcally, we ﬁnd support for an inverse relationship between explainability and accuracy from the perspective of cost experts. We also ﬁnd that the accurate but complex and less explainable ML model of gradient boosted regression (GBR) was preferred in only a few situations; mostly, the more basic, better explainable models of multiple linear regression (MLR) and case-based reasoning (CBR) were preferred, although these were less accurate. This suggests that lack of explainability can indeed be a major limitation for the application of ML models. Furthermore, we investigate speciﬁc characteristics that could increase task uncertainty and the importance of explainability in our context: project unpredictability, product cost granularity, predecessor product availability, target cost gap, and product development phase. Keywords: Machine learning; Explainability; Cost estimation; Task uncertainty; 1. Introduction Machine learning receives increasing attention in management accounting (Abernethy et al., 2023; Amani & Fadlalla, 2017; Fehrenbacher et al., 2023; Kuzey et al., 2019; LaValle et al., 2011; Losbichler & Lehner, 2021; Mahlendorf et al., 2023; Nielsen, 2018; Repenning et al., 2022; Rikhardsson & Yigitbasioglu, 2018; Sutton et al., 2016) . However, the applicability of complex machine learning (ML) models is often limited by the a

FAQs

How does task uncertainty affect the value of explainability compared with accuracy?

Higher task uncertainty increases the desire to interact with the ML model, raising the relative importance of explainability over accuracy.

What relationship between explainability and accuracy is supported by the cost experts’ perspective?

The study finds support for an inverse relationship between explainability and accuracy from the perspective of cost experts.

Which ML models were preferred and why does this matter for real applications?

Despite gradient boosted regression (GBR) being accurate but complex and less explainable, it was preferred only in a few situations; more basic and better-explainable models such as multiple linear regression (MLR) and case-based reasoning (CBR) were often preferred even though they were less accurate.

What factors in the paper increase task uncertainty in the product cost estimation context?

The paper identifies project unpredictability, product cost granularity, predecessor product availability, target cost gap, and the product development phase as characteristics that increase task uncertainty and thus the importance of explainability.

Aug 5, 2026

Upload by Clementine