Search results

Gender bias in artificial intelligence (AI) should be solved as a priority before AI algorithms become ubiquitous, perpetuating and accentuating the bias. While the problem has been identified as an established research and policy agenda, a cohesive review of existing research specifically addressing gender bias from a socio-technical viewpoint is lacking. Thus, the purpose of this study is to determine the social causes and consequences of, and proposed solutions to, gender bias in AI algorithms.

A comprehensive systematic review followed established protocols to ensure accurate and verifiable identification of suitable articles. The process revealed 177 articles in the socio-technical framework, with 64 articles selected for in-depth analysis.

Most previous research has focused on technical rather than social causes, consequences and solutions to AI bias. From a social perspective, gender bias in AI algorithms can be attributed equally to algorithmic design and training datasets. Social consequences are wide-ranging, with amplification of existing bias the most common at 28%. Social solutions were concentrated on algorithmic design, specifically improving diversity in AI development teams (30%), increasing awareness (23%), human-in-the-loop (23%) and integrating ethics into the design process (21%).

This systematic review is the first of its kind to focus on gender bias in AI algorithms from a social perspective within a socio-technical framework. Identification of key causes and consequences of bias and the breakdown of potential solutions provides direction for future research and policy within the growing field of AI ethics.

The peer review history for this article is available at https://publons.com/publon/10.1108/OIR-08-2021-0452

Managerial mindset and cognitive bias can be barriers to any transformation strategy. In the case of telework, most employees express willingness to telework, yet, few firms formally enable it during regular business hours. The status quo is a daily commute to the traditional workplace. The purpose of this paper is to test framing interventions designed to harness cognitive biases through choice architecture.

Drawing upon behavioral strategy and prospect theory, this paper presents two studies: quasi-experiments with 146 senior business students and experiments in the field (replication using random assignment and extension) with 84 senior decision makers. Both studies use a one-way between-subjects design and chi-square analysis.

Findings support the proposition that, although cognitive biases can act as barriers to transformation, they can be re-framed through strategic interventions. Specifically, in both studies, there was a drastic increase in adoption simply by changing the way the choice was presented. Findings in the lab were cross-validated in the field. Observed shifts in preferences provide evidence that embedding the right reference point within communications can frame a decision choice more favorably. Findings also support that a bias for an implicitly perceived status quo can be overruled through an explicitly stated reference point.

It is an assumption of behavioral strategy that most individuals simply respond to the gains/loss framing without being influenced by other psychological or contextual factors, and though these effects dissipate through aggregation, it is a limitation nonetheless. Indeed, using an individual construct to explain an organizational phenomenon is a well-debated topic in the field of strategy, with proponents on both sides. The distinguishing factor, here, is that behavioral strategists are only interested in results at the aggregated level.

Practitioners attempting to roll out telework adoption, or any transformation, now have proven strategies for designing frames of reference that intervene against and harness the power of loss aversion and the status quo.

This paper measures micro processes that have an effect at the macro level. It explains systematic aversion to adoption as an aggregation of decision-making behavior that is seemingly subconscious. In doing so, it highlights the impact of bounded rationality perpetuated through social systems, while measuring effective interventions designed to make systematic behavior more predictable.

A novel contribution is made in designing/testing a new frame for systematic resistance to change that frames the status quo as the losing prospect. In this frame, the perceived loss is in the choice not to change, and loss aversion proves to be an effective tool for facilitating systematic change.

This paper presents a mathematical programming model to reduce bias for both aggregate demand forecasts and lower echelon forecasts comprising a hierarchical forecasting system. Demand data from an actual service operation are used to illustrate the model and compare its accuracy with a standard approach for hierarchical forecasting. Results show that the proposed methodology outperforms the standard approach.

Online reviews in online health communities (OHCs) have been a vital information source for patients. The extant literature on the bias effects of helpful reviews mainly concentrates on traditional e-commerce, whereas research on OHCs is still rare. Thus, based on the heuristic-systematic model (HSM), this research explores how two unique reviewer characteristics in OHCs, which may induce attribution bias and confirmation bias, affect review helpfulness and how review length moderates these relationships.

This research analyzed 130,279 reviews collected from haodf.com (one of the representative OHCs in China) by adopting the negative binomial regression to test our research model.

The results indicate that reviewer cured status positively influences review helpfulness, whereas reviewer recommendation source negatively affects review helpfulness. Moreover, the effects of the two reviewer cues on review helpfulness will be weaker for longer reviews.

First, as one of the initial attempts, the current study investigates the effects of confirmation bias and attribution bias of online reviews in OHCs by exploring the effects of two unique reviewer characteristics on review helpfulness. Second, the weakening moderating effects of review length on the two bias effects provide empirical support for the theoretical arguments of the HSM in OHCs.

This article aims to systematically review the literature published in recognized journals focused on cognitive heuristic-driven biases and their effect on investment management activities and market efficiency. It also includes some of the research work on the origins and foundations of behavioral finance, and how this has grown substantially to become an established and particular subject of study in its own right. The study also aims to provide future direction to the researchers working in this field.

For doing research synthesis, a systematic literature review (SLR) approach was applied considering research studies published within the time period, i.e. 1970–2021. This study attempted to accomplish a critical review of 176 studies out of 256 studies identified, which were published in reputable journals to synthesize the existing literature in the behavioral finance domain-related explicitly to cognitive heuristic-driven biases and their effect on investment management activities and market efficiency as well as on the origins and foundations of behavioral finance.

This review reveals that investors often use cognitive heuristics to reduce the risk of losses in uncertain situations, but that leads to errors in judgment; as a result, investors make irrational decisions, which may cause the market to overreact or underreact – in both situations, the market becomes inefficient. Overall, the literature demonstrates that there is currently no consensus on the usefulness of cognitive heuristics in the context of investment management activities and market efficiency. Therefore, a lack of consensus about this topic suggests that further studies may bring relevant contributions to the literature. Based on the gaps analysis, three major categories of gaps, namely theoretical and methodological gaps, and contextual gaps, are found, where research is needed.

The skillful understanding and knowledge of the cognitive heuristic-driven biases will help the investors, financial institutions and policymakers to overcome the adverse effect of these behavioral biases in the stock market. This article provides a detailed explanation of cognitive heuristic-driven biases and their influence on investment management activities and market efficiency, which could be very useful for finance practitioners, such as an investor who plays at the stock exchange, a portfolio manager, a financial strategist/advisor in an investment firm, a financial planner, an investment banker, a trader/broker at the stock exchange or a financial analyst. But most importantly, the term also includes all those persons who manage corporate entities and are responsible for making their financial management strategies.

Currently, no recent study exists, which reviews and evaluates the empirical research on cognitive heuristic-driven biases displayed by investors. The current study is original in discussing the role of cognitive heuristic-driven biases in investment management activities and market efficiency as well as the history and foundations of behavioral finance by means of research synthesis. This paper is useful to researchers, academicians, policymakers and those working in the area of behavioral finance in understanding the role that cognitive heuristic plays in investment management activities and market efficiency.

Survey-based studies suggest that search engines are trusted more than social media or even traditional news, although cases of false information or defamation are known. The purpose of this paper is to analyze query suggestion features of three search engines to see if these features introduce some bias into the query and search process that might compromise this trust. The authors test the approach on person-related search suggestions by querying the names of politicians from the German Bundestag before the German federal election of 2017.

This study introduces a framework to systematically examine and automatically analyze the varieties in different query suggestions for person names offered by major search engines. To test the framework, the authors collected data from the Google, Bing and DuckDuckGo query suggestion APIs over a period of four months for 629 different names of German politicians. The suggestions were clustered and statistically analyzed with regards to different biases, like gender, party or age and with regards to the stability of the suggestions over time.

By using the framework, the authors located three semantic clusters within the data set: suggestions related to politics and economics, location information and personal and other miscellaneous topics. Among other effects, the results of the analysis show a small bias in the form that male politicians receive slightly fewer suggestions on “personal and misc” topics. The stability analysis of the suggested terms over time shows that some suggestions are prevalent most of the time, while other suggestions fluctuate more often.

This study proposes a novel framework to automatically identify biases in web search engine query suggestions for person-related searches. Applying this framework on a set of person-related query suggestions shows first insights into the influence search engines can have on the query process of users that seek out information on politicians.

Purpose – A major limitation in cross-cultural research continues to be attempts to compare construct measurements across cultures without adequate conceptual and empirical evidence of the equivalency of the measurement scores. Of significant concern in such studies is the presence of various types of response bias that may systematically differ from one culture to another, resulting in a potential violation of the assumption that measurement scores across cultures are equivalent. The focus of this study is to investigate the role of the response format type, extreme response style (ERS). Most studies have investigated response bias styles using Likert-type scales as response formats, yet it has long been argued that these particular formats tend to result in various types of response style bias, especially in cross-cultural research. Would other scaling devices, such as semantic differential (SD), lessen response style bias in pan-cultural studies? To answer this question, our study employs two types of response formats (i.e., Liker-type and SD) to empirically test for the presence of ERS within each response format style.

Methodology/approach – This chapter takes the form of empirical research using ERS indices to test for the degree of ERS between response formats using samples from a collectivistic culture (i.e., South Korea) and an individualistic culture (i.e., United States).

Findings – Results show that samples from both cultures exhibit greater levels of ERS when using Likert-type scales compared to SD scales. Additionally, this study finds that, when using Likert-type scales, ERS is greater for U.S. respondents than for South Korea respondents. Finally, results show that there is no statistically significant difference in ERS between the two cultural groups when using SD response formats.

Research implications – Findings show that the use of SD response formats eliminates systematic differences in ERS between a collectivist sample and an individualist sample. Therefore, the use of such response formats in future cross-cultural research can greatly diminish the problematic effects of culturally based ERS and lead to greater confidence in the equivalency of measurement scores across cultures.

Originality/value of paper – This study is the first to simultaneously assess culturally based ERS using two types of response formats to investigate the impact of response format on ERS. Furthermore, this study assesses the role of response format on ERS both within and between two distinctly different cultures.

This study extends prior accounting research on decision aids (DAs) relating to face validity. Specifically, this study aims to examine the effects of face validity through the presence of two levels of bias in DA output. The presence of bias in a DA will not affect how statistically informative an aid is but will decrease the face validity. The findings suggest that non-expert DA users recognize the bias in the DA’s suggestions as evidenced by users’ low agreement with the aid; however, they do not adjust for the bias in their performance, suggesting that non-expert users do not learn from the DA.

This repeated-measures experimental design allows us to examine performance effects over time in response to different levels of bias in the DA output. The participants in the study are provided with outcome feedback to examine learning effects.

The findings suggest that non-expert DA users recognize the bias in the DA’s suggestions as evidenced by users’ low agreement with the aid; however, they do not adjust for the bias in their performance, suggesting that non-expert users do not learn from the DA. Although users of an unbiased DA strongly agree with the DA’s output, individual performance deteriorates over time. Initially, the users of an unbiased DA perform better than those who use a biased DA; however, over time, the performance of users of an unbiased aid deteriorates and the performance of users of the biased aid does not improve.

Companies developing DAs may need to consider the effects of using a DA under circumstances different from those under which the aid was developed and that may lead to the biased DA output. This study has implications for firms that design, develop and use DAs.

This study considers a yet unexamined face validity issue – observable bias in DA output. This study examines deterministic DAs designed to assist the decision-maker through their ability to combine multiple cues in a systematic and consistent manner. This study has implications for firms that design, develop and use DAs. Firms need to consider the effects of using a DA under circumstances different from those under which the aid is developed, thereby, potentially leading to biased DA output. Each additional variable added to the DA will be associated with an incremental cost in a DA’s development, use and modification. The results of this study provide insights contributing to the information available for cost–benefit analyses conducted when developing a DA or when considering the modification of existing aid. Failure to change a DA because of face validity issues alone may result in a decline in user performance. Thus, the cost of modifying a DA must be weighed against the benefits resulting from improved performance. This study contributes insights into how users’ responses to DA bias could affect the assessments of the benefits of including an omitted variable in a DA.

This article aims to clarify the mechanism by which heuristic-driven biases influence the entrepreneurial strategic decision-making in an emerging economy.

Entrepreneurs' heuristic-driven biases have been measured using a questionnaire, comprising numerous items, including indicators of entrepreneurial strategic decision-making. To examine the relationship between heuristic-driven biases and entrepreneurial strategic decision-making process, a 5-point Likert scale questionnaire has been used to collect data from the sample of 169 entrepreneurs who operate in small- and medium-sized enterprises (SMEs). The collected data were analyzed using SPSS and Amos graphics software. Hypotheses were tested using structural equation modeling (SEM) technique.

The article provides empirical insights into the relationship between heuristic-driven biases and entrepreneurial strategic decision-making. The results suggest that heuristic-driven biases (anchoring and adjustment, representativeness, availability and overconfidence) have a markedly negative influence on the strategic decisions made by entrepreneurs in emerging markets. It means that heuristic-driven biases can impair the quality of the entrepreneurial strategic decision-making process.

The article encourages entrepreneurs to avoid relying on cognitive heuristics or their feelings when making strategic decisions. It provides awareness and understanding of heuristic-driven biases in entrepreneurial strategic decisions, which could be very useful for business actors such as entrepreneurs, managers and entire organizations. Understanding regarding the role of heuristic-driven biases in entrepreneurial strategic decisions may help entrepreneurs to improve the quality of their decision-making. They can improve the quality of their decision-making by recognizing their behavioral biases and errors of judgment, to which we are all prone, resulting in a more appropriate selection of entrepreneurial opportunities.

The current study is the first to focus on links between heuristic-driven bias and the entrepreneurial strategic decision-making in Pakistan—an emerging economy. This article enhanced the understanding of the role that heuristic-driven bias plays in the entrepreneurial strategic decisions and more importantly, it went some way toward enhancing understanding of behavioral aspects and their influence on entrepreneurial strategic decision-making in an emerging market. It also adds to the literature in the area of entrepreneurial management specifically the role of heuristics in entrepreneurial strategic decision-making; this field is in its initial stage, even in developed countries, while, in developing countries, little work has been done.

Indecision is not the hallmark of a great manager, but what do we know about when and why managers avoid or postpone decisions? The purpose of this paper is to discuss the limited research on indecision.

This article reviews research from judgment and decision making, psychology, management, and marketing literatures to assemble what we already know about indecision. The review spans situational and personal determinants of indecision, highlighting what we know about when and why people experience indecision as well as who is predictably indecisive.

Decisions are avoided when people are asked to justify them, when options are similar in attractiveness, and when there are a large number of options to consider. Indecision may sometimes be a result of systematic biases (i.e. omission bias and status quo bias), and indecisive people may be more prone to confirmation bias. Finally, indecisiveness is related to numerous other individual differences, many of which are negative.

Specific recommendations for managers include evaluating options separately rather than comparing options, structuring incentive systems to reward decisive action, and explicitly considering the risk of lost opportunity when deciding whether to put off making a decision.

The literature reviewed in this paper spans diverse disciplines and perspectives. This paper provides a starting point for managers and researchers interested in understanding indecision: when and why it occurs, who is likely to be most indecisive, and what we might do to counter indecision.