Search results

This paper aims to investigate whether professional data analysts’ trust of black-box systems is increased by explainability artifacts.

The study was developed in two phases. First a black-box prediction model was estimated using artificial neural networks, and local explainability artifacts were estimated using local interpretable model-agnostic explanations (LIME) algorithms. In the second phase, the model and explainability outcomes were presented to a sample of data analysts from the financial market and their trust of the models was measured. Finally, interviews were conducted in order to understand their perceptions regarding black-box models.

The data suggest that users’ trust of black-box systems is high and explainability artifacts do not influence this behavior. The interviews reveal that the nature and complexity of the problem a black-box model addresses influences the users’ perceptions, trust being reduced in situations that represent a threat (e.g. autonomous cars). Concerns about the models’ ethics were also mentioned by the interviewees.

The study considered a small sample of professional analysts from the financial market, which traditionally employs data analysis techniques for credit and risk analysis. Research with personnel in other sectors might reveal different perceptions.

Other studies regarding trust in black-box models and explainability artifacts have focused on ordinary users, with little or no knowledge of data analysis. The present research focuses on expert users, which provides a different perspective and shows that, for them, trust is related to the quality of data and the nature of the problem being solved, as well as the practical consequences. Explanation of the algorithm mechanics itself is not significantly relevant.

Perception has been identified as the main cause underlying most autonomous vehicle related accidents. As the key technology in perception, deep learning (DL) based computer vision models are generally considered to be black boxes due to poor interpretability. These have exacerbated user distrust and further forestalled their widespread deployment in practical usage. This paper aims to develop explainable DL models for autonomous driving by jointly predicting potential driving actions with corresponding explanations. The explainable DL models can not only boost user trust in autonomy but also serve as a diagnostic approach to identify any model deficiencies or limitations during the system development phase.

This paper proposes an explainable end-to-end autonomous driving system based on “Transformer,” a state-of-the-art self-attention (SA) based model. The model maps visual features from images collected by onboard cameras to guide potential driving actions with corresponding explanations, and aims to achieve soft attention over the image’s global features.

The results demonstrate the efficacy of the proposed model as it exhibits superior performance (in terms of correct prediction of actions and explanations) compared to the benchmark model by a significant margin with much lower computational cost on a public data set (BDD-OIA). From the ablation studies, the proposed SA module also outperforms other attention mechanisms in feature fusion and can generate meaningful representations for downstream prediction.

In the contexts of situational awareness and driver assistance, the proposed model can perform as a driving alarm system for both human-driven vehicles and autonomous vehicles because it is capable of quickly understanding/characterizing the environment and identifying any infeasible driving actions. In addition, the extra explanation head of the proposed model provides an extra channel for sanity checks to guarantee that the model learns the ideal causal relationships. This provision is critical in the development of autonomous systems.

Research into the interpretability and explainability of data analytics and artificial intelligence (AI) systems is on the rise. However, most recent studies either solely promote the benefits of explainability or criticize it due to its counterproductive effects. This study addresses this polarized space and aims to identify opposing effects of the explainability of AI and the tensions between them and propose how to manage this tension to optimize AI system performance and trustworthiness.

The author systematically reviews the literature and synthesizes it using a contingency theory lens to develop a framework for managing the opposing effects of AI explainability.

The author finds five opposing effects of explainability: comprehensibility, conduct, confidentiality, completeness and confidence in AI (5Cs). The author also proposes six perspectives on managing the tensions between the 5Cs: pragmatism in explanation, contextualization of the explanation, cohabitation of human agency and AI agency, metrics and standardization, regulatory and ethical principles, and other emerging solutions (i.e. AI enveloping, blockchain and AI fuzzy systems).

As in other systematic literature review studies, the results are limited by the content of the selected papers.

The findings show how AI owners and developers can manage tensions between profitability, prediction accuracy and system performance via visibility, accountability and maintaining the “social goodness” of AI. The results guide practitioners in developing metrics and standards for AI explainability, with the context of AI operation as the focus.

This study addresses polarized beliefs amongst scholars and practitioners about the benefits of AI explainability versus its counterproductive effects. It poses that there is no single best way to maximize AI explainability. Instead, the co-existence of enabling and constraining effects must be managed.

The author augments an otherwise standard business-cycle model with a rich government sector and adds monopolistic competition in the product market and rigid prices, as well as rigid wages a la Calvo (1983) in the labor market.

This specification with the nominal wage rigidity, when calibrated to Bulgarian data after the introduction of the currency board (1999–2018), allows the framework to reproduce better observed variability and correlations among model variables and those characterizing the labor market in particular.

As nominal wage frictions are incorporated, the variables become more persistent, especially output, capital stock, investment and consumption, which help the model match data better, as compared to a setup without rigidities.

The findings suggest that technology shocks seem to be the dominant source of economic fluctuations, but nominal wage rigidities as well as the monopolistic competition in the product market, might be important factors of relevance to the labor market dynamics in Bulgaria, and such imperfections should be incorporated in any model that studies cyclical movements in employment and wages.

The computational experiments performed in this paper suggest that wage rigidities are a quantitatively important model ingredient, which should be taken into consideration when analyzing the effects of different policies in Bulgaria, which is a novel result.

In times of organizational thirst for employee engagement and meaning through designing corporate stories, the aim of this article is to explore and identify key sources (engines) of engagement during LEGO® SERIOUS PLAY® (LSP) corporate learning pre-pandemic events of various types and size in Poland.

This is a conceptual paper. The research was conducted using participant observation from the perspective of a certified facilitator of the method. This position ensures a prime access to the organizational events. Eight training sessions (four LSP and four non-LSP workshops) have been analysed using thematic analysis. The structure of thematic codes has been conceptualized and reflected as the EPIC framework.

The findings include (1) the importance of the experience of emerging realities as a key generator of engagement, (2) the significance of social collaboration and peer-to-peer interactions (experience of collective intelligence), (3) the observable rise in engagement and willingness to contribute when real business situations, especially labelled as “strategic issues” are discussed and (4) the role of image-capturing (“snapshot experience”) in creation of an engaging learning experience.

The limitations refer to the potential conflict of interests as the researcher is also the facilitator of the workshop. To ensure the neutral point of view of the researcher, the sessions have been recorded to enable transparency of the observation and non-biased logic of key findings. The “learning experience” research is also culture- and context-sensitive, thus it may be problematic to replicate the research procedure in different countries, however, the EPIC model can be treated as a universal framework to explore and identify the engines of engagement.

The concept of this paper is designed from the practical point of view. The findings are adaptable to the corporate practices aimed at empowering employees and are compatible with management models such as agile, human enablement and human-centred design in organizations.

Serious play methods of learning and experiencing are said to be of the highest importance when finding new ways of organizational learning in the pandemic situation and work from home as a standard learning environment.

The contribution of this paper is visible in the conceptualization of the moments that shape an engaging experience. This is also the first academic paper presenting the perspective of a certified facilitator of LSP from Central and Eastern Europe region.

This study aims to investigate whether firms purchase directors' and officers' liability (D&O) insurance when the country-level economic policy uncertainty (EPU) is high.

This study uses D&O insurance data from Chinese listed firms between 2003 and 2019 to conduct regression analyses to examine the association between D&O insurance and EPU.

The results show that government EPU, despite being an exogenous factor, increases the likelihood of firms' purchasing D&O insurance, and this effect is more pronounced when firms are exposed to great share price crash risk and high litigation risk, suggesting that firms intend to purchase D&O insurance possibly due to the accentuated stock price crash risk and litigation risk associated with EPU. In addition, the results indicate that the effect of EPU on the D&O insurance purchase decision is moderated by the provincial capital market development and internal control quality.

The study highlights the role of uncertain economic policies in shareholder approval of D&O insurance purchases.

The study enriches the literature on the determinants of D&O insurance purchases by documenting novel evidence that country-level EPU is a key institutional factor shaping firms' decisions to purchase D&O insurance.

This study aims to depict how a change laboratory (CL) promotes sustainable professional practice at the workplace to tackle unequal access to educational success.

The empirical findings are from a CL focusing on school professionals’ agency and a follow-up study one year after the CL.

The study shows how the staff gained insight that professional agency is a collective and relational practice. Furthermore, the staff explored how to make a difference with viable means to create new workplace models for students’ success despite experiencing a conundrum.

This study examined participants’ perspectives in workplace change and provided support for further research examining how professionally and collectively designed models gain sustainability in schools.

This study provides empirical data of how professional agency for change driven by collective visions can be accelerated with the interventionist method CL among school professionals.

This study emphasizes the value of professional collective learning at the workplace, driven by several professional groups in school, and the need to follow up to detect sustainable change.

This study emphasizes the value of professional collective learning at the workplace, driven by several professional groups in school, and the need to follow up to detect sustainable change.

In the rapidly growing e-grocery segment, unattended delivery is an emerging practice with the potential to offer a superior delivery experience. The purpose of this study is to contribute to the body of knowledge for unattended grocery delivery services by empirically identifying and describing the forms and determinants of customer expectations.

A multiple case study of potential early adopters was conducted to explore customer expectations of unattended grocery delivery services. Empirical data collected from direct observations and semi-structured interviews with ten Swedish households were coded and put through a single-case as well as a cross-case analysis revealing emerging patterns from which propositions were formed.

The iteration of theory and data in the case study resulted in a conceptual model of service expectations and determinants, containing six propositions. The study reveals a clear pattern that consumers expect to save time, gain flexibility and benefit from the ease of use of the service, while they predict sufficient security. Moreover, consumers’ desire open access features from retailers and service providers, integrated product returns service and nondescript hardware designs. The findings suggest that these service expectations are determined by personal needs, technology literacy and situational factors. The identified personal needs are stress reduction, limiting social interaction and increasing spare time.

To support further theory development, this study presents six propositions for the types, forms and determinants of customer expectations of unattended grocery delivery.

This study provides managers with up-to-date insights into customer expectations and offers guidance in designing and developing unattended grocery delivery services.

This study contains the first in-depth analysis of customer expectations of unattended grocery delivery services, which are increasingly used for last mile e-grocery delivery.

This study aims to examine how leadership identity manifests at the individual and collective levels within a relational training context among a group of multicultural higher education students.

This is a case study and examines the interactions among eight multicultural students through the theoretical lens of leadership identity development (LID) theory.

The main findings of this study suggest that LID manifests through an open will and intensifying motivation to the collective impulse of achieving shared goals through nurturing the collective cognition to integrate diverse perspectives and a broadening view of leadership as a collective capacity for co-creation and generativity.

Although the paper builds on a case study with a limited number of participants and the ability to generalise its findings is partial, the study may provide practical applications for training leadership in other collaborative contexts and supporting it at the individual and collective levels.

The LID theory and LID model have been applied simultaneously to a training lab to examine how LID manifests among a multicultural group of higher education students. The lab emphasises a participatory leadership-oriented pedagogy.