Fairness is an important topic that has captured the interest of researchers in many fields. Looking at behavioural and experimental economics, various methods have been used to shed the light on fairness. One of the most recognizable ways is through the use of the ultimatum game. In this paper, we aimed to look at fairness considerations by utilizing this game, while also highlighting gender and culture as factors of importance. Two models, specifically Rabin’s reciprocity model and Fehr and Schmidt’s inequity aversion model, in addition to various research papers on the topics of gender and culture, were used as the cornerstones for this paper. The experimental design was a replication of the ultimatum game, which was carried out online. Results showed that there were noticeable effects from both gender and culture on the behaviour of the participants in the game. However, the results were not statistically significant enough to be considered as determining factors in regard to why individuals behave in a specific manner in the game. Since research from the field has reached mixed conclusions, we can say that making inferences about human behaviour is harder than one might think.

<div>The past few years have seen an increasing focus on fairness and the long-term impact of algorithmic decision making in the context of Machine learning, Artificial Intelligence and other disciplines. In this thesis, we model hiring processes in enterprises and organizations using dynamic mechanism design. Using a stochastic block model to simulate the workings of a hiring process, we study fairness and long-term evolution in the system. </div><div> </div><div> We first present multiple results on a deterministic variant of our model including convergence and an accurate approximate solution describing the state of the deterministic variant after any time period has elapsed. Using the differential equation method, it can be shown that this deterministic variant is in turn an accurate approximation of the evolution of our stochastic block model with high probability.</div><div> </div><div> Finally, we derive upper and lower bounds on the expected state at each time step, and further show that in the limiting case of the long-term, these upper and lower bounds themselves converge to the state evolution of the deterministic system. These results offer conclusions on the long-term behavior of our model, thereby allowing reasoning on how fairness in organizations could be achieved. We conclude that without sufficient, systematic incentives, under-represented groups will wane out from organizations over time.</div>

QUIC is a relatively new transport layer network protocol that has gained popularity over the last few years. The protocol was initially developed by Google and standardization work has been continued by the Internet Engineering Task Force (IETF) with the goal of it becoming the next generation transport protocol. While the standardization work is not yet finished, the protocol has seen a large adoption, already covering a large portion of the internet traffic. As a new protocol, many researchers have studied QUIC and compared it to TCP in typical scenarios. However, few studies have been performed on QUIC in specific scenarios. In this thesis, we present the first performance study of QUIC over Dual Connectivity (DC). DC is a multi-connectivity technique that allows users to connect to multiple cell towers with one user equipment. It is an important lower-layer feature accelerating the transition from 4G to 5G, which is also expected to play an important role in standalone 5G networks. With DC, higher throughput and reliability can be achieved by using multiple paths simultaneously. However, the drawback of DC is that it introduces packet reordering and jitter, which can significantly impact the performance of upper-layer protocols such as TCP and QUIC. To study the extent of this effect, a testbed is set up to evaluate QUIC over DC. Our performance evaluation compares the throughput of QUIC over DC with that of TCP over DC, and evaluates the fairness of QUIC over DC. Using a series of throughput and fairness experiments, we show how QUIC is affected by different DC parameters, network conditions, and whether the DC implementation aims to improve throughput or reliability. Our findings provide network operators with insights into understanding the impacts of splitting QUIC traffic in a DC environment. We show the value of increasing the UDP receive buffers when running QUIC over DC and that QUIC can utilize the increased bandwidth and reliability in DC, provided that the links' characteristics are similar. We also show that with reasonably selected DC parameters and increased UDP receive buffers, QUIC over DC performs similarly to TCP over DC and achieves optimal systemwide fairness under symmetric link conditions when DC is not used for packet duplication.

The service sector is the fastest growing segment of the economy, responsible for 75% of the GNP, and still growing. Its success is important to the global economy. Nonetheless, throughout the 20-year evolution of services marketing literature, research that guides theory, methodology, and practice for service success has remained underrepresented. Published research regarding the effect of customers' justice perceptions on customer satisfaction is primarily experimental and focuses only on service recovery after a service failure, providing insufficient information about how the justice experienced in a service encounter affects a customer's satisfaction level. Proactive and reactive service recovery research abounds; service failures have overshadowed service success. This is the first empirical research to investigate across service outcomes the effects 1) of interactional, distributive, and procedural justice on overall justice and customer satisfaction and 2) of overall justice on customer satisfaction. The theoretical model of the customer's path to loyalty adapts previous models of the service profit chain, customer satisfaction with service failure and recovery, and complaint handling relationships. It is a simplified version of the author's in-work conceptual model. The theoretical model has conceptual and practical value to researchers and service company executives. It considers all possible service encounter types and the heterogeneity of outcomes. It is supported by attribution and equity theories (the underpinnings of customer's justice judgments) and by behavioral intentions research. A cross-sectional written survey was used to gather data relevant to the eight hypotheses proposed and shown on the measurement model. Sixty percent of the 302 respondents recalled satisfying service encounters and 40% recalled dissatisfying service encounters. MANOVA testing supported the hypothesis of a positive relationship for extant prior experience to each of the justice constructs. The tested path analysis model showed direct and positive effects for the justice constructs on overall justice and customer satisfaction and for overall justice on customer satisfaction. When providers fairly address the people, outputs, and processes in service transactions, expectations are more likely to be met, delight is possible, and trust and commitment, possibly even loyalty, may arise. Disappointment and disconfirmation resulting from gaps in performance expectations can lead to non-attritive defection and lost profits. This research provides practical information that can lead to a better understanding of customers' evaluation methods and be used to guide the formation of improved service strategies that provide justice, a key to satisfaction. / Ph. D.

Machine learning algorithms are increasingly used by decision making systems that affect individual lives in a wide variety of ways. Consequently, in recent years concerns have been raised about the social and ethical implications of using such algorithms. Particular concerns include issues surrounding privacy, fairness, and transparency in decision systems. This dissertation introduces new tools and measures for improving the social desirability of data-driven decision systems, and consists of two main parts. The first part provides a useful tool for an important class of decision making algorithms: collaborative filtering in recommender systems. In particular, it introduces the idea of improving socially relevant properties of a recommender system by augmenting the input with additional training data, an approach which is inspired by prior work on data poisoning attacks and adapts them to generate `antidote data' for social good. We provide an algorithmic framework for this strategy and show that it can efficiently improve the polarization and fairness metrics of factorization-based recommender systems. In the second part, we focus on fairness notions that incorporate data inputs used by decision systems. In particular, we draw attention to `data minimization', an existing principle in data protection regulations that restricts a system to use the minimal information that is necessary for performing the task at hand. First, we propose an operationalization for this principle that is based on classification accuracy, and we show how a natural dependence of accuracy on data inputs can be expressed as a trade-off between fair-inputs and fair-outputs. Next, we address the problem of auditing black- box prediction models for data minimization compliance. For this problem, we suggest a metric for data minimization that is based on model instability under simple imputations, and we extend its applicability from a finite sample model to a distributional setting by introducing a probabilistic data minimization guarantee. Finally, assuming limited system queries, we formulate the problem of allocating a query budget to simple imputations for investigating model instability as a multi-armed bandit framework, for which we design efficient exploration strategies.

Neuropsychological assessment in linguistically heterogeneous populations is fraught with numerous challenges, such as lacking or inappropriate normative data or the unavailability of appropriate tests. Accommodating multilingual individuals exacerbates the issue by adding the question of which language(s) to use when assessing multilingual individuals. Different testrelated concepts may be accessible to them via different languages, as their lexicon is spread out over two or more languages. Hence, any monolingual instrument is likely to disadvantage them. The present set of three studies circumvents this question and presents evidence for an inherently multilingual English/Afrikaans/isiXhosa screening tool for intelligence, the Multilingual Vocabulary Test (MVT). I describe the instrument’s development from the pilot study to a psychometric analysis of the final, digitally administered version. For an abbreviated 13-item version, Study 3 (N = 494) shows an internal consistency of  = .59 and Study 2 (N = 101) produced significant criterion-related validity values of r = .46 and r = .52 with the KBIT-2 and Shipley-2 VIQ scores respectively. Linear regression analyses show that, while all criterion measures are biased toward E1-speakers, the MVT is largely immune to test-takers’ linguistic background. Thus, the MVT paves the way toward more fairness in cognitive assessments, in general, and provides a promising first step toward addressing one of South African neuropsychologists’ greatest needs—that of a quick and easy-to-administer, yet linguistically fair screening tool for cognitive impairment.

We studied perceptions parental differential treatment as reported by parents and children in two different settings. Perceptions of differential affection and control were examined in healthy families and in families that include a child diagnosed with Type 1 diabetes. Parental differential treatment was assessed using questionnaires that measured perceptions of absolute parenting for children and their siblings. Difference scores were subsequently utilized to generate perceived parental differential treatment scores. Participants were 61 parents (half with healthy children, half with one child who has diabetes) and 62 children (half comprising sibling pairs unaffected by any medical problems, half including one child with Type 1 diabetes). Children within the sibling pairs were between 11 and 18 years of age and approximately two years apart, on average. Parents were also asked about their children's emotional/behavioral adjustment and adherence to prescribed medical regimen (in the diabetes group), and their levels of parenting stress. Children were also administered measures regarding their emotional/behavioral adjustment, average adherence (in the diabetes group), and perceptions of deservedness of parental treatment perceived. No differences in strength of correlations between ratings of parental differential treatment and child adjustment iv were detected across groups. Significant differences, however, emerged with regard to type of perceived parental differential treatment that related to child adjustment scores across groups. Relationships were also detected between perceived parental differential treatment and ratings of adherence and measures of glycemic control in the diabetes group. Perceived deservedness as rated by children, ratings of absolute parenting, and parenting stress were observed to moderate the relationship between ratings of parental differential treatment and child adjustment. Parental differential treatment scores predicted unique variance in reported child behavior problems above and beyond that predicted by absolute parenting measures. Differences in relationships across groups, the role of gender, and the importance of context and family in studying perceptions of parental differential treatment and child adjustment are discussed.

京都大学 / 0048 / 新制・課程博士 / 博士(経済学) / 甲第20654号 / 経博第554号 / 新制||経||282(附属図書館) / 京都大学大学院経済学研究科経済学専攻 / (主査)教授 小佐野 広, 教授 柴田 章久, 准教授 敦賀 貴之 / 学位規則第4条第1項該当 / Doctor of Economics / Kyoto University / DGAM

Diversity and inclusivity in the business world has gained larger focus from media and customers over the past 60 years, however a large focus has been aimed at gender, ethnicity and race whilst other types of diversity have not been as researched. The inclusivity of people with disabilities in the regular labor market has been neglected to a large extent. All people should be included in the workplace which can be seen as a reflection of society as a whole, especially in regard to the Swedish model of government agency assistance from the company Samhall. In this study we have combined frameworks from diversity management, critical disability theory and justice theory to create a study examining how the developmental opportunities of people with disabilities are affected by assumptions and fairness. This study was conducted with the help of Samhall's regional office in Västerbotten and Örnsköldsvik. Samhall employs 26 000 people in Sweden with the goal of creating meaningful work for people with reduced working ability, developing their skills to be ready for the regular labor market. This study was designed from a managerial perspective, and we therefore decided to use a qualitative approach by interviewing six managers from Samhall using a semi-structured format. We wanted to gain insight on how society and other companies interact with Samhall, and how they work with development to reach their final goals. Due to the limited number of studies about people with disabilities from a social standpoint we were able to establish a research gap. Many of the current studies from a social perspective are focused on developmental perspective from young ages, or the perspective is from a medical standpoint. Studies with people with disabilities as a part of diversity management have yet to be created to any greater extent. By further including a perspective of equality and justice in this we were able to address issues of conflict as well, as a part of developmental work. The interviewees from Samhall are managers with varying amounts of experience within Samhall and varying experiences working with people with disabilities before their work in the organization. The managers also had varied job assignments which were able to contribute to a more diverse perspective on everything from daily interactions with society to policy creation and medical paperwork. In our study we researched six different assumptions about disabilities including the social model of disability, valuing diversity, rights, voices, language and transformative politics. From diversity management we used instrumental and terminal values to help us understand the developmental work. Finally, we used a framework of equality categorization including equality rule, relative justice and objective justice to compare how things work within and outside of Samhall and see what the employees respond to in the best way and to further cultivate development for the employees. We did a thematic analysis to be able to compare the answers of our different interviewees and find out what really works and to see what the current situations truly look like. The research showed that assumptions do affect people with disabilities, however, not to the same extent as previous literature suggests. Furthermore, we discovered that the instrumental values used in the developmental work for people with disabilities in Samhall can be transferred to other companies as well, and therefore hopefully intrigue and incentivize more employers to take the step of hiring from Samhall and people with disabilities in general. The fairness aspect of our research also provided insight in how to prioritize and how this can aiding and resolving conflict.

No description available.