No abstract

No abstract

Background Dermoscopy is commonly used for the evaluation of pigmented lesions, but agreement between experts for identification of dermoscopic structures is known to be relatively poor. Expert labeling of medical data is a bottleneck in the development of machine learning (ML) tools, and crowdsourcing has been demonstrated as a cost- and time-efficient method for the annotation of medical images. Objective The aim of this study is to demonstrate that crowdsourcing can be used to label basic dermoscopic structures from images of pigmented lesions with similar reliability to a group of experts. Methods First, we obtained labels of 248 images of melanocytic lesions with 31 dermoscopic “subfeatures” labeled by 20 dermoscopy experts. These were then collapsed into 6 dermoscopic “superfeatures” based on structural similarity, due to low interrater reliability (IRR): dots, globules, lines, network structures, regression structures, and vessels. These images were then used as the gold standard for the crowd study. The commercial platform DiagnosUs was used to obtain annotations from a nonexpert crowd for the presence or absence of the 6 superfeatures in each of the 248 images. We replicated this methodology with a group of 7 dermatologists to allow direct comparison with the nonexpert crowd. The Cohen κ value was used to measure agreement across raters. Results In total, we obtained 139,731 ratings of the 6 dermoscopic superfeatures from the crowd. There was relatively lower agreement for the identification of dots and globules (the median κ values were 0.526 and 0.395, respectively), whereas network structures and vessels showed the highest agreement (the median κ values were 0.581 and 0.798, respectively). This pattern was also seen among the expert raters, who had median κ values of 0.483 and 0.517 for dots and globules, respectively, and 0.758 and 0.790 for network structures and vessels. The median κ values between nonexperts and thresholded average–expert readers were 0.709 for dots, 0.719 for globules, 0.714 for lines, 0.838 for network structures, 0.818 for regression structures, and 0.728 for vessels. Conclusions This study confirmed that IRR for different dermoscopic features varied among a group of experts; a similar pattern was observed in a nonexpert crowd. There was good or excellent agreement for each of the 6 superfeatures between the crowd and the experts, highlighting the similar reliability of the crowd for labeling dermoscopic images. This confirms the feasibility and dependability of using crowdsourcing as a scalable solution to annotate large sets of dermoscopic images, with several potential clinical and educational applications, including the development of novel, explainable ML tools.

Scholars, politicians, and laypeople alike bemoan the high level of political polarization in the United States, but little is known about how to bring the views of liberals and conservatives closer together. Previous research finds that providing people with information regarding a contentious issue is ineffective for reducing polarization because people process such information in a biased manner. Here, we show that information can reduce political polarization below baseline levels and also that its capacity to do so is sensitive to contextual factors that make one’s relevant preferences salient. Specifically, in a nationally representative sample (Study 1) and a preregistered replication (Study 2), we find that providing a taxpayer receipt—an impartial, objective breakdown of how one’s taxes are spent that is published annually by the White House—reduces polarization regarding taxes, but not when participants are also asked to indicate how they would prefer their taxes be spent.

Purpose Advances in information technology have enabled new ways of organizing work and led to a proliferation of what is known as the “gig economy.” While much attention has been paid to how these new organizational designs have upended traditional employee–employer relationships, there has been little consideration of how these changes have impacted the social norms and expectations that govern the relationship between workers and consumers. The purpose of this paper is to consider the social norm of tipping and propose that gig work is associated with a breakdown of tipping norms in part because of workers’ increased autonomy in terms of deciding when and whether to work. Design/methodology/approach The authors present four studies to support their hypothesis: a survey vignette experiment with workers on Amazon Mechanical Turk (Study 1), an analysis of New York City taxi data (Study 2), a field experiment with restaurant employee food delivery drivers (Study 3) and a field experiment with gig-worker food delivery drivers (Study 4). Findings In Studies 1 and 2, they find that consumers are less likely to tip when workers have autonomy in deciding whether to complete a task. In Study 3, they find that restaurant delivery employees notice upfront tips (or lack thereof) and alter their service as a result. In contrast, in Study 4, they find that gig-workers who agree to complete a delivery for a fixed amount that includes an upfront tip (or lack thereof) are not responsive to tips. Together, these findings suggest that the gig economy has not only transformed employee-employer relationships, but has also altered the norms and expectations of consumers and workers. Originality/value The authors present four different studies that consider the social norm of tipping in the context of gig work. Together, they highlight that perceptions of worker autonomy have driven the decline in tipping norms associated with gig work.