Developmental and evolutionary constraints on olfactory circuit selection.

SignificanceIn this work, we explore the hypothesis that biological neural networks optimize their architecture, through evolution, for learning. We study early olfactory circuits of mammals and insects, which have relatively similar structure but a huge diversity in size. We approximate these circuits as three-layer networks and estimate, analytically, the scaling of the optimal hidden-layer size with input-layer size. We find that both longevity and information in the genome constrain the hidden-layer size, so a range of allometric scalings is possible. However, the experimentally observed allometric scalings in mammals and insects are consistent with biologically plausible values. This analysis should pave the way for a deeper understanding of both biological and artificial networks.

One model for the learning of language.

A major goal of linguistics and cognitive science is to understand what class of learning systems can acquire natural language. Until recently, the computational requirements of language have been used to argue that learning is impossible without a highly constrained hypothesis space. Here, we describe a learning system that is maximally unconstrained, operating over the space of all computations, and is able to acquire many of the key structures present in natural language from positive evidence alone. We demonstrate this by providing the same learning model with data from 74 distinct formal languages which have been argued to capture key features of language, have been studied in experimental work, or come from an interesting complexity class. The model is able to successfully induce the latent system generating the observed strings from small amounts of evidence in almost all cases, including for regular (e.g., an , [Formula: see text], and [Formula: see text]), context-free (e.g., [Formula: see text], and [Formula: see text]), and context-sensitive (e.g., [Formula: see text], and xx) languages, as well as for many languages studied in learning experiments. These results show that relatively small amounts of positive evidence can support learning of rich classes of generative computations over structures. The model provides an idealized learning setup upon which additional cognitive constraints and biases can be formalized.

The acquired dyad inclination and decreased interpersonal brain communication in the pursuit of collective benefit.

People perform better collectively than individually, a phenomenon known as the collective benefit. To pursue the benefit, they may learn from previous behaviors, come to know whose initial opinion should be valued, and develop the inclination to take it as the collective one. Such learning may affect interpersonal brain communication. To test these hypotheses, this study recruited participant dyads to conduct a perceptual task on which they made individual decisions first and then the collective one. The enhanced interpersonal brain synchronization (IBS) between participants was explored when individual decisions were in disagreement vs. agreement. Computational modeling revealed that participant dyads developed the dyad inclination of taking the higher-able participants', not the lower-able ones' decisions as their collective ones. Brain analyses unveiled the enhanced IBS at frontopolar areas, premotor areas, supramarginal gyri, and right temporal-parietal junctions. The premotor IBS correlated negatively with dyad inclination and collective benefit in the absence of correction. The Granger causality analyses further supported the negative relation of dyad inclination with inter-brain communication. This study highlights that dyads learn to weigh individuals' decisions, resulting in dyad inclinations, and explores associated inter-brain communication, offering insights into the dynamics of collective decision-making.

Going beyond the comparison: toward experimental instructional design research with impact.

To design effective instruction, educators need to know what design strategies are generally effective and why these strategies work, based on the mechanisms through which they operate. Experimental comparison studies, which compare one instructional design against another, can generate much needed evidence in support of effective design strategies. However, experimental comparison studies are often not equipped to generate evidence regarding the mechanisms through which strategies operate. Therefore, simply conducting experimental comparison studies may not provide educators with all the information they need to design more effective instruction. To generate evidence for the what and the why of design strategies, we advocate for researchers to conduct experimental comparison studies that include mediation or moderation analyses, which can illuminate the mechanisms through which design strategies operate. The purpose of this article is to provide a conceptual overview of mediation and moderation analyses for researchers who conduct experimental comparison studies in instructional design. While these statistical techniques add complexity to study design and analysis, they hold great promise for providing educators with more powerful information upon which to base their instructional design decisions. Using two real-world examples from our own work, we describe the structure of mediation and moderation analyses, emphasizing the need to control for confounding even in the context of experimental studies. We also discuss the importance of using learning theories to help identify mediating or moderating variables to test.

Medical students' motivations for participating in an elective focused on social inequalities and health disparities.

In this study, we examine students' reasons for pursuing elective training focused on medical racism and systemic health inequities at a midwestern medical school. Data collection included semi-structured interviews with students who participated in an optional course focused on these topics. We analyzed their motivations, goals, and interests using reflexive thematic analysis and created three themes based on students' responses. Theme (1) "pre-existing conditions" focuses on students' knowledge, beliefs, worldviews and experience prior to the class. Theme (2) "enacting change" examines their desires to become effective physicians and improve medicine overall. Theme (3) "creating community" considers their preferences for a supportive and connected learning and social environment. We discuss the findings within the context of adult learning theory and Self-Determination Theory. The research provides insight about the overt and underlying factors that drive medical students' participation in training focused on social inequality. We also share recommendations for curriculum development and future research based on the patterns we found in students' discussions of their needs and expectations.

Public health education using social learning theory: a systematic scoping review.

BACKGROUND: Public health education (PHE) in social environments plays a crucial role in mitigating the impact of public health events, especially with the recent surge in global incidents. Social learning theory (SLT) provides a strong theoretical foundation for implementing PHE. The objective of this study is to conduct a systematic scoping review of PHE using SLT, synthesizing the target populations, types of research, main findings, and future directions. METHODS: The study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Extension for Scoping Review (PRISMA-ScR) guidelines. We conducted a comprehensive search of five electronic databases (Web of Science, Scopus, PubMed, ProQuest, and APA PsycInfo) for English articles related to PHE using SLT. Two reviewers independently screened the titles and abstracts. Descriptive statistics were utilized to analyze the characteristics of the articles included in the study, followed by a comprehensive narrative analysis of the results. RESULTS: Research on PHE using SLT mainly focuses on adolescents, students, special patients, and vulnerable populations. The study sample includes seven research types and nine commonly used experimental methods. Four modes of PHE using SLT are identified, along with four types of summarized research results. CONCLUSION: PHE research based on SLT can be prioritized for preventing widespread infectious diseases, spreading fundamental public health information, and assisting patients with particular illnesses. To enhance the implementation of PHE, researchers and policymakers should integrate online and offline health education resources, ensure the accessibility of up-to-date information, and leverage digital technologies in PHE. More highly interactive and participatory health education courses will be established in social learning environments to encourage public participation in PHE.

Belief polarization in a complex world: A learning theory perspective.

We present two models of how people form beliefs that are based on machine learning theory. We illustrate how these models give insight into observed human phenomena by showing how polarized beliefs can arise even when people are exposed to almost identical sources of information. In our first model, people form beliefs that are deterministic functions that best fit their past data (training sets). In that model, their inability to form probabilistic beliefs can lead people to have opposing views even if their data are drawn from distributions that only slightly disagree. In the second model, people pay a cost that is increasing in the complexity of the function that represents their beliefs. In this second model, even with large training sets drawn from exactly the same distribution, agents can disagree substantially because they simplify the world along different dimensions. We discuss what these models of belief formation suggest for improving people's accuracy and agreement.

Short-Term Effects of an eHealth Care Experiential Learning Program Among Patients With Type 2 Diabetes: Randomized Controlled Trial.

BACKGROUND: Type 2 diabetes is a chronic disease with a significant medical burden. eHealth care integrates medicine and technology to enhance the outcomes of such patients; however, adequate eHealth literacy (eHL) is necessary for that to happen. Fostering eHL is crucial for patients with diabetes to engage with eHealth care and receive quality care and timely support. Experiential learning theory can enhance patients' eHL and skills to use eHealth care technology in their daily care. OBJECTIVE: This study explored the effectiveness of an eHealth care experiential learning program in improving eHL, patient health engagement, and eHealth care use status among patients with type 2 diabetes in 3 months. METHODS: In this randomized controlled trial, patients under case management services from various clinics in Taiwan were randomly assigned to either the intervention group receiving the 6-session eHealth care experiential learning program or the control group receiving the usual care. Data were collected using structured questionnaires at 3 time points: pretest, postintervention, and 3 months after the intervention. Descriptive data were presented using frequency distribution, percentage, mean, and SD. The outcomes were analyzed using a generalized estimating equation method by intention-to-treat analysis. RESULTS: A total of 92 participants (46 in each group) were recruited in this study. Of these, 86 completed the course and follow-up evaluations with a mean age of 62.38 (SD 12.91) years. After completing the intervention, the intervention group had significantly higher posttest scores in eHL (ß=19.94, SE 3.52; P<.001), patient health engagement (ß=.28, SE 0.13; P=.04), and eHealth use (ß=3.96, SE 0.42; P<.001) than the control group. Furthermore, the intervention group maintained these significant improvements in eHL (ß=18.19, SE 3.82; P<.001) and eHealth use (ß=3.87, SE 0.49; P<.001) after 3 months. CONCLUSIONS: Participating in the eHealth care experiential learning program resulted in significant improvements in eHL, patient health engagement, and eHealth use among patients with type 2 diabetes. Our interventional program can inform future clinical practice and policies to strengthen self-management skills and facilitate the use of health technology in caring for patients with chronic diseases. TRIAL REGISTRATION: ClinicalTrials.gov NCT05180604; https://clinicaltrials.gov/ct2/show/NCT05180604.

Interpersonal communication and perceived norms as social influence mechanisms of e-cigarette use among adults: a systematic review.

Background: Given the increasing popularity of e-cigarette use among adults and the ongoing debate about the benefits and the potential adverse health risks associated with e-cigarette use, it is critical to identify the correlates of e-cigarette use. Prior research has found associations between interpersonal communication, perceived norms, and adults' e-cigarette use, but the evidence has yet to be summarized and synthesized.Objectives: This paper reviewed empirical studies examining the relationship between interpersonal communication, perceived norms, and e-cigarette use among adults.Methods: Following PRISMA guidelines, articles were searched on DOAJ, EMBASE, Europe PubMed Central, Google Scholar, PsychINFO, PubMed, Web of Science, and the reference list of the retrieved studies for studies that examined social influence on e-cigarette use. Three reviewers independently screened 1,713 non-duplicate papers and further screened the full text of 195 articles for inclusion.Results: Thirty studies (30), consisting of quantitative (n = 25) and qualitative (n = 5) data, were included in this review. The twenty-five (25) quantitative studies consisted of both cross-sectional (n = 20) and longitudinal (n = 5) studies. Interpersonal communication portraying e-cigarettes as beneficial or harmful was found to increase e-cigarette use and quit attempts, respectively. Across study designs, greater perceptions of others' e-cigarette use or approval were related to more frequent e-cigarette use.Conclusions: The findings highlight that e-cigarette-related interpersonal communication and perceived norms are associated with e-cigarette use. These factors may be useful targets in brief interventions. However, most of the included studies were cross-sectional, limiting the ability to establish clear cause-and-effect relationships; therefore, more longitudinal studies are needed.

Longitudinal variation of correlations between different components of assessment within a medical school.

BACKGROUND: An assessment program should be inclusive and ensure that the various components of medical knowledge, clinical skills, and professionalism are assessed. The level and the variation over time in the strength of the correlation between these components of assessment is still a matter of study. Based on the meaningful learning theory and the integrated learning theory, we hypothesize that these components increase their connections during the medical school course. METHODS: This is a retrospective cohort study that analyzed data collected for a 10-year period in one medical school. We included students from the 3rd to 6th year of medical school from 2011 to 2021. Three assessment components were addressed: Knowledge, Clinical Skills, and Professionalism. For data analysis, Pearson correlation coefficients (R) and R2 were calculated to study the correlation between variables and a z-test on Fisher's r-to-z was used to determine the differences between correlation coefficients. RESULTS: 949 medical students were included in the study. The correlation between Clinical Skills and Professionalism showed a medium to strong association (Pearson's R ranging from 0.485 to 0.734), while the correlation between Knowledge and Professionalism was weaker but exhibited a steady evolution with Pearson's R fluctuating between 0.075 and 0.218. The Knowledge and Clinical Skills correlation became statistically significant from 2013 onwards and peaking at Pearson's R of 0.440 for the cohort spanning 2016-2019. We also revealed a strengthening of correlations between Professionalism and Clinical Skills from the beginning to the end of clinical training, but not with the knowledge component. CONCLUSIONS: This analysis contributes to our understanding of the dynamics of correlations of different assessment components within an institution and provides a framework for how they interact and influence each other. TRIAL REGISTRATION: This study was not a clinical trial, but a retrospective observational study, without health care interventions. Nevertheless, we provide herein the number of the study as submitted to the Ethics committee - CEICVS 146/2021.

Professionalism-training in undergraduate medical education in a multi-cultural, multi-ethnic setting in the Gulf Region: an exploration of reflective essays.

BACKGROUND: Despite the established need to prioritize professionalism-training in developing future physicians, very few medical programs in the Gulf Region embed in their curricula discrete contextualized courses aimed at developing the corresponding competencies, while fostering self-directed learning. This study aims at exploring the perception of undergraduate medical students in a multi-cultural, multi-ethnic setting regarding their understanding of, and personal experience with professionalism through their engagement with the content of an innovative curriculum-based professionalism course, offered at a Medical School in Dubai, United Arab Emirates. METHODS: The study used a qualitative phenomenological research design. Out of 33 students, 29 students had submitted reflective essays. The content of these essays was inductively analyzed following a six-step framework for conducting thematic analysis. The framework's steps include familiarizing oneself with the data, generating initial codes, searching for themes, reviewing themes, defining and naming themes, and producing the report. FINDINGS: The inductive qualitative analysis generated the Professionalism Learning Journey model. This conceptual model includes four interconnected themes: Awareness, Acknowledgement, Realization, and Application. The generated model depicts the trajectory that the learners appear to experience while they are engaging with the content of the course. CONCLUSION: Integrating a professionalism-training course into an undergraduate medical curriculum is likely to be positively appraised by the learners. It raises their awareness, enables them to value the subject matter and the sophistication of its application, and empowers them to put into practice the taught principles, on an individual basis and collectively. This is especially true when the course is entrenched in constructivism experiential learning theory and designed to foster self-directed learning. The introduced conceptual model, in conjunction with the innovative professionalism-training course curriculum, can serve as a template for other competencies and other schools.

A spark of change: developing an innovative gerontological nursing intervention mapping initiative for training and education (IGNITE).

BACKGROUND: With an aging global population and advancements in medical technology, there is an urgent need for innovative gerontological nursing education programs. This study aimed to develop and evaluate the Innovative Gerontological Nursing Intervention Mapping Initiative for Training and Education (IGNITE) program. This program is a digital platform-based postgraduate nursing curriculum that employs the Intervention Mapping Approach (IMA) and Transformative Learning Theory to address the evolving needs of gerontological nursing. METHODS: The IGNITE program's development process encompassed a comprehensive approach, including needs assessment, mapping of course objectives, integration of theory-based methods and strategies, course design, implementation, and rigorous evaluation. The pilot evaluation study involved pre- and post-tests focused on ageism, attitudes towards elder care, knowledge about older adults, transformative behavior change, and program satisfaction. The findings revealed significant improvements across all these dimensions, affirming the effectiveness of the program. RESULTS: The program leveraged experiential learning, critical reflection, and rational discourse to facilitate transformative educational experiences. Notably, pre- and post-test comparisons showed marked improvements in attitudes towards older adult care and dementia care knowledge. Participants expressed high satisfaction with the program, with significant reported changes in transformative behaviors. The study also illuminated the initial negative attitudes of clinical nurses towards older adults and underscored the importance of transformative learning experiences in fostering empathy and understanding. CONCLUSIONS: The IGNITE program lays a foundational framework for developing educational materials that promote transformative learning and self-reflection among healthcare professionals. This approach can lead to innovative nursing practices and personal growth. The application of the IMA and Transformative Learning Theory in gerontological nursing education shows significant promise. Future research should focus on exploring the long-term impacts of such programs and their applicability in diverse healthcare settings.

How do mandatory emergency medicine rotations contribute to the junior residents' professional identity formation: a qualitative study.

BACKGROUND: This study aims to investigate the impact of short-term mandatory emergency medicine rotations on professional identity formation of Japanese junior residents. Using situated learning theory as a theoretical framework, we explore how this rotation, which is part of a two-year Junior residency in the transition period from students to qualified physicians. METHODS: We conducted a qualitative study conducting semi-structured face-to-face interviews with Year 1 postgraduate residents in the 2020-2021 classes of the junior residency program in Okinawa Chubu Hospital, Japan (n = 10). The data obtained from the interviews were analysed using inductive thematic analysis to identify the themes regarding professional identity formation. RESULTS: Four main themes regarding professional identity formation emerged from the data analysis: patient care, teamwork, role models, and peers. Junior residents said they had the opportunity to participate in the emergency department community and experience training in authentic clinical contexts. Clinical exposure influenced the professional identity formation of the junior residents. Nurses and peers played a crucial role in this. Junior residents see the training in the emergency department as the beginning of their careers. CONCLUSION: Short-term mandatory rotations enabled junior residents to integrate into the emergency department community, demonstrating autonomy and responsibility. These experiences fostered their professional identity by helping their socialisation within the community of practice.

Becoming a doctor: using social constructivism and situated learning to understand the clinical clerkship experiences of undergraduate medical students.

BACKGROUND: Despite the emphasis on the uniqueness and educational importance of clinical clerkships in medical education, there is a lack of deep understanding of their educational process and outcomes. Especially due to an inherent trait of clinical clerkships which requires participation in the workplace outside the classroom, it is difficult to fully comprehend their educational potential using traditional learning perspectives such as imbibing outside knowledge. Accordingly, this study aims to explore the experiences of a rotation-based clerkship of medical school students from the perspective of social constructivism of learning, which can empirically examine what and how medical students learn during clinical clerkship in South Korea. By providing an insight into the workings of the clerkship process, this study contributes to a better understanding of how a learning-friendly environment can be cultivated at clinical clerkships. METHODS: The study utilized a basic qualitative study to understand what and how medical students learn during their clinical clerkships. Semi-structured, in-depth individual interviews were conducted with eight sixth-graders who had experienced a two-year clerkship at Ajou University Medical School. Data were analyzed based on Lave and Wenger's situated learning theory and Wenger's social theory in learning. RESULTS: We found that the medical students had developed different aspects of their professional identities such as values, functionality, career decisions, sociality, and situating during their clinical clerkships. Further, professional identity was formed through a combination of participation and reification-the processes involved in the negotiation of meaning. This combination was facilitated by the students' first experience and relationships with professors, classmates, and patients. Finally, non-learning occurred in the context of over-participation (learning anxiety and alienation) or over-reification (evaluation and e-portfolio). CONCLUSIONS: This study revealed five sub-professional identities and their formation process from the learners' perspective, thereby uncovering the unique learning characteristics and advantages of rotated-based clerkship and contributing to a further understanding of how gradual improvements can be made to the traditional clerkship education of medical students.

Traditional social learning predicts cyber deviance? Exploring the offending versatility thesis in social learning theory.

Social learning theory has been widely implemented to understand cyber deviance. Nevertheless, the antecedent scholarship homogenously nested in the perspective of offending specification, leaving the offending versatility thesis unattained. The lack of such studies may undermine the capability of comprehensively understanding the social learning patterns of online offending. Using a sample of 3741 Chinese college students, this study estimated an array of binary logistic regressions to compare the effects of traditional and online social learning in four types of online offending (online sexual harassment, cyberbullying, hacking, and digital piracy). The results suggest that offending versatility and offending specification co-exist in the social learning process of cyber deviance, while offending specification explains a marginally greater variance. Besides, online learning variables act as potential mediators in the relationships between traditional learning and cyber deviance. Furthermore, traditional social learning shows greater predictive power in cyber-enabled crimes than in cyber-dependent crimes. Our study provides fresh empirical evidence for the non-exclusive association between offending versatility and offending specification in the social learning process of cyber deviance.

Examining the effectiveness of education based on social learning theory in fostering self-care and social skills in school children: A randomized controlled trial.

PURPOSE: This study was conducted to examine the effectiveness of a training program prepared on the basis of the Social Learning Theory for teaching self-care and social competence behaviors in preschool children. DESIGN AND METHOD: The sample of the study consisted of 41 in the intervention, 41 in the control group. Data were collected with Demographic Assessment Form, the Self-Care Skills Assessment Scale and the Social Competence and Behavior Evaluation-30 Scale. Using five animated films created on the basis of the Social Learning Theory for fostering children self-care and social competence behaviors. RESULTS: The scores of the children in the intervention group on the three sub-dimensions of the self-care scale and the total score of the scale were found to be significantly higher than the control group (p < 0.05). The intervention group scored significantly higher on the self-care scale and social competence and behavior scales than the control group (p < 0.05). CONCLUSION: As a result of the research, it was seen that the applied training significantly affected the self-care and social competence behaviors in the intervention group. It was seen that using Social Learning Theory in teaching health behaviors has been effective in the learning process of preschool children. IMPLICATIONS TO PRACTICE: In preschool period, children learn by observing and imitating their adopted role models. Since, we suggest that health training program should be based on the stages of Social Learning Theory. Animated films that focus on the target health behaviors can be used as training tools.

"It´s like walking in a bubble", nursing students´ perspectives on age suit simulation in a home environment - group interviews from reflection seminars.

BACKGROUND: Older persons with age-related and complex health problems will increasingly depend on care provision from nurses in their own homes. However, a barrier to quality care is ageism and nursing students´ disinterest in geriatrics. In addition, nurse education often falls short in preparing students for the complexity of geriatric care. Welfare technology (WT) is progressively implemented in home care to help older persons live at home despite their health problems. However, this process is intricate and requires acceptance and digital literacy among caregivers and older persons. Despite these challenges, nurse education can address and change negative attitudes through innovative teaching methods such as age suit simulation. Therefore, the study aims to describe nursing students´ experiences of age suit simulation in a home-like environment with WT and technical aids, and will reveal their perspective on ageing and providing care to older adults. METHODS: A qualitative explorative design using semi-structured group interviews (n=39) among nursing students. Data was analysed through reflexive thematic analysis. RESULTS: The analysis generated three main themes; "It's like walking in a bubble", "An eye opener" and "Concerns about ageing and the current structure of geriatric care". The main themes included eight subthemes. Adapting to the sensory and physical limitations of the age suit was an immersive experience and caused feelings of frustration, loneliness and disconnection. A prominent result was a raised awareness of cognitive loss, especially impaired vision, and students felt the simulations had made them aware of the everyday challenges older persons faced. Students highlighted the importance of patience and giving enough time in care situations by being present and having a critical perspective of WT. The students were mostly negative towards their own ageing and could better relate to older persons´ vulnerability. CONCLUSIONS: Age suit simulation was described as an embodied and eye-opening experience, raising nursing students´ awareness of older persons´ functional limitations and the consequences for dignity and independence. Coping with cognitive loss was especially difficult. Students were motivated to apply their new knowledge to clinical practice. Age suit simulation can complement geriatric education, preparing students for the complex care needs of older persons.

Impact of a Grand Rounds Interprofessional Workshop: student perceptions of interprofessional socialization and cultural humility.

While uniprofessional education programs develop strong student identities, they may limit the development of behaviors needed for interprofessional socialization. Interprofessional education (IPE) creates an essential platform for student engagement in the development of interprofessional socialization and cultural humility, thus enabling improvement in collaborative communication. In this quasi-experimental observational study, health professional students attended one of three Grand Rounds Interprofessional Workshops (GRIW) and completed online pre- and post-workshop surveys including sociodemographic background, the Interprofessional Socialization and Valuing Scale (ISVS), and the Cultural Competence Self-Assessment Checklist (CCSAC). A total of 394 students from eight professions participated in the workshop with 287 (73%) of attendees completing both pre- and post-workshop surveys. No significant differences were observed in ISVS and CCSAC scores between students across workshops. Significant pre- to post-workshop differences were found in ISVS [t (284) = 13.5, p < .001, 95%], CCSAC [t (286) = 13.8, p < .001] and the cultural competence components of cultural awareness [t (285) = 12.9, p < .001, 95%], knowledge [t (285) = 9.5, p < .001, 95%], and skills [t (286) = 13.3, p < .001, 95%]. Interprofessional education learning opportunities that integrate socialization with health professional students and cultural humility education can improve educational awareness of cultural values and communication for collaborative professional practice.

From fringe to mainstream: The Garcia effect.

The rejection of research results is sometimes thought to be justified in cases of individuals embracing fringe ideas that depart significantly from prevailing orthodoxy, or in cases of individuals who lack appropriate expertise or credentials. The case of John Garcia exhibits both of these dimensions, and illustrates that such rejection can delay scientific advancements. Garcia's work decisively challenged what was the orthodoxy in psychology in the midcentury: behaviorism. Behaviorist learning theorists suffered from theory-entrenchment insofar as they failed to acknowledge Garcia's anomalous research findings that ran counter to their theoretical expectations. The case study also illustrates that theories on the margins can become embraced as a result of advancements in adjacent research fields. Studying how Garcia's work moved from fringe to mainstream results in lessons for the philosophy of science and epistemology more generally. Only when we see the mechanisms of exclusion at work can we understand how science and other knowledge production systems can inadvertently act counterproductively via gatekeeping practices that filter out unorthodox points of view.

Benign overfitting in linear regression.

The phenomenon of benign overfitting is one of the key mysteries uncovered by deep learning methodology: deep neural networks seem to predict well, even with a perfect fit to noisy training data. Motivated by this phenomenon, we consider when a perfect fit to training data in linear regression is compatible with accurate prediction. We give a characterization of linear regression problems for which the minimum norm interpolating prediction rule has near-optimal prediction accuracy. The characterization is in terms of two notions of the effective rank of the data covariance. It shows that overparameterization is essential for benign overfitting in this setting: the number of directions in parameter space that are unimportant for prediction must significantly exceed the sample size. By studying examples of data covariance properties that this characterization shows are required for benign overfitting, we find an important role for finite-dimensional data: the accuracy of the minimum norm interpolating prediction rule approaches the best possible accuracy for a much narrower range of properties of the data distribution when the data lie in an infinite-dimensional space vs. when the data lie in a finite-dimensional space with dimension that grows faster than the sample size.