Computing students often encounter real-world problems that are open-ended, ambiguous, and lack a single correct solution. This paper examines how computing students navigate such problems in collaborative team settings and how these experiences shape peer evaluation. Using data from a project-based visualization course, we analyze team Discord discussions, participation and activity metrics, and written peer evaluations across multiple teams through a mixed-methods approach. Our findings show that teams commonly encounter decision conflict when working on open-ended problems, which they address through negotiation, resource sharing, and iterative discussion. Disagreement often coincides with sustained engagement and collaborative sensemaking while avoiding breakdown. Instructors are typically treated as collaborative resources, with teams retaining ownership over decisions. Peer evaluations reflect these processes only weakly: visible participation and time investment show limited association with peer ratings, while peers emphasize the relevance, timing, and impact of contributions. Differences between self- and peer evaluations further highlight the socially situated nature of peer judgment. These findings provide empirical insight into collaborative open-ended problem solving in computing education and how peer evaluation captures students' perceptions of contribution.

Hackathons have emerged as dynamic platforms for fostering innovation, collaboration, and skill development in the technology sector. Structural differences across hackathon formats raise important questions about how event design can shape student learning experiences and engagement. This study examines two distinct hackathon formats: a gender-specific hackathon (GS) and a regular institutional hackathon (RI). Using a mixed-methods approach, we analyze variations in team dynamics, project themes, role assignments, and environmental settings. Our findings indicate that GS hackathon foster a collaborative and supportive atmosphere, emphasizing personal growth and community learning, with projects often centered on health and well-being. In contrast, RI hackathon tend to promote a competitive, outcome-driven environment, with projects frequently addressing entertainment and environmental sustainability. Based on these insights, we propose a hybrid hackathon model that combines the strengths of both formats to balance competition with inclusivity. This work contributes to the design of more engaging, equitable, and pedagogically effective hackathon experiences.

Children go through several developmental milestones at an early age. When a child becomes slow or cannot even reach one or more of these milestones compared to other typically developing children, that condition is called Developmental Delays and Disabilities (DDDs). Children's DDDs create much parental stress, resulting in social isolation and loneliness. YouTubers make videos on DDDs to assist parents and other caregivers. We need to gain more knowledge of how these videos are created and how much they impact the parents of children with DDDs. We followed a mixed-method approach to annotate and analyze 1,532 YouTube videos on children's DDDs. Our analysis found that YouTube videos on DDDs support the community with critical informational content and provide mental and emotional strength through content related to personal experiences. Our analysis of the comments in these videos explored how YouTubers and viewers created a strong community connection through these video content. Most importantly, the interviews with the parents of children with DDDs revealed how parents find these videos relatable and critical for taking necessary actions for their children's DDDs diagnosis and follow-up treatments. We concluded by discussing platform-centric design implications for supporting parents and other caregivers of children with DDDs.

Over the past 30 years, the enrollment of students of color in higher education has increased significantly because of various efforts taken nationwide for promoting diversity and equity. Yet, graduation rates for these students have remained stagnant. This research explores the multifaceted challenges encountered by underrepresented students in STEM disciplines, focusing on Black and Latinx communities. The study, involving students predominantly from an East Coast R1 research university, employs a mixed-methods approach, utilizing interviews, surveys, and focus groups to examine the student of color experience within the contexts of sense of belonging, role models, and Diversity, Equity, and Inclusion (DEI) initiatives, in addition to general challenges. Thematic analysis facilitated comprehensive data interpretation. Our findings underscore the pressure to persist despite doubts, the significance of in-group connections, and differing experiences based on intersecting identities. The absence of diverse faculty emerged as a poignant concern, influencing students' sense of belonging. Mentorship and role models were identified as critical components for fostering a supportive environment. Regarding DEI initiatives, participants called for more robust efforts emphasizing the need for dedicated spaces to address concerns such as microaggressions. This research provides a comprehensive exploration of the challenges faced by underrepresented STEM students and offers actionable insights for enhancing inclusivity. The findings emphasize the need for targeted interventions, including increased faculty diversity, structured mentorship programs, and improvements in DEI initiatives. Implementing these recommendations is crucial for fostering an environment where every student can thrive and contribute to the rich tapestry of STEM disciplines.

Legal crowdfunding is an emerging domain where lawyers and individuals raise funds to fight legal actions. To study how prospective donors can verify the credibility of legal campaigns, we analyzed the conversations surrounding these campaigns on Facebook. We discovered three primary themes associated with the perceptions of the contributors of legal campaigns: supporters posting admiring and appreciative comments, supporters posting critical and disapproving comments, and opponents posting critical and disapproving comments. We observed that while supporters criticized campaigns' opponents, biased media, and dishonest authorities, opponents criticized campaign owners, campaigns' objectives and opaque logistics. To understand the impact of these perspectives on donors, we followed up with an online survey study where we presented a legal campaign with its corresponding social media conversations. We found that critical comments impacted donation decisions more than appreciative comments. We concluded with design implications to better support potential donors to make more informed donation decisions.

The COVID-19 pandemic has affected more than 301 million people worldwide so far. Many communities (such as minority communities) suffered disproportionately more difficulties throughout the pandemic. In this paper, we would like to focus on one such community: COVID-19 long-haulers community. Long-hauler community consists of people affected by Coronavirus, but their symptoms do not cure in a couple of weeks; instead, they experience lingering symptoms for months. The concerns of this community were initially ignored by health care providers primarily because of limited information. In this paper, we have analyzed the social media discussion of a private Facebook group dedicated to the long-hauler community. In addition, we interviewed the community members to investigate their motivations for joining the group and how the group has impacted their lives as long-hauler patients. Our analyses revealed the primary discussion topics of this community. It also showed how a minority community could stand by each other using social media groups during a crisis. We concluded the paper with long-term implications of our findings for health care systems, policies, and existing literature on cooperative AI.

India produces a large amount of shrimp every year and exports them globally. Yet, concerns have been raised about shrimps produced in India not qualifying the global safety standards. In this paper, we took an ethnographic approach to understand the traditional farming practices followed by shrimp farmers in India. We interviewed 29 shrimp farmers from Andhra Pradesh in India for eight months. We learned that most experienced and large-scale shrimp farmers in India prefer to follow traditional practices in their shrimp farms. They prefer to adopt technology-driven solutions in their farming only if absolutely needed. However, we observed that young, small-scale farmers are more open to trying new technologies in their farming. We discussed several areas where initiatives can be taken to encourage more Indian shrimp farmers to adopt new technology-driven solutions in their farming.

In recent years, political crowdfunding campaigns have emerged through which politicians raise money to fund their election campaigns. Divisive issues discussed in these campaigns may not only motivate donations but also could have a broader priming effect on people’s social opinions. In the U.S., more than one-third of the population with moderate opinions show a tendency to swing their opinion based on recent and more accessible events. In this paper, we ask: can such campaigns further prime people’s responses to partisan topics, even when we discuss those topics in a non-political context? To answer this question, we analyzed the influence of exposure to a political candidate’s crowdfunding campaign on responses to a subsequently seen, unrelated scientific topic that is not inherently political but is seen as partisan in the U.S. (climate change). We found that exposure to an attitude-inconsistent political candidate’s crowdfunding campaign (a campaign that is counter to someone’s existing political beliefs) can have a significant priming effect on subsequently seen politically charged topics. This effect may occur due to the activation of in-group identity by the candidate’s partisan campaign. Guided by these findings, we investigated elements that can mitigate this self-categorization effect. We found that carefully designed content following framing techniques such as schema framing and threat/safety framing can mitigate people’s sense of self-categorization toward non-political topics.

Markov Logic Networks (MLNs) represent relational knowledge using a combination of first-order logic and probabilistic models. In this paper, we develop an approach to explain the results of probabilistic inference in MLNs. Unlike approaches such as LIME and SHAP that explain black-box classifiers, explaining M LN inference is harder since the data is interconnected. We develop an explanation framework that computes importance weights for MLN formulas based on their influence on the marginal likelihood. However, it turns out that computing these importance weights exactly is a hard problem and even approximate sampling methods are unreliable when the MLN is large resulting in non-interpretable explanations. Therefore, we develop an approach where we reduce the large MLN into simpler coalitions of formulas that approximately preserve relational dependencies and generate explanations based on these coalitions. We then weight explanations from different coalitions and combine them into a single explanation. Our experiments illustrate that our approach generates more interpretable explanations in several text processing problems as compared to other state-of-the-art methods.

People encounter video ads every day when they access online content. While ads can be annoying or greeted with resistance, they can also be seen as informative and enjoyable. We asked the question, what might make an ad more enjoyable? And, do people with different personality traits prefer to watch different ads --- could it be possible to better match ads and people? To answer these questions, we conducted an online study where we asked people to watch video ads of different emotional sentiments. We also measured their personality traits through an online survey. We found that the sentiment of people's preferred video ads varies significantly based on their personality traits. Additionally, we investigated when these ads are accompanied by content, how the emotional state induced by accompanying content affects people's ad preferences. We found that there was a complex relationship between people's emotional state induced by accompanying content and their ad preference when an ad highlighted either an alertness or calmness sentiment. However, when an ad highlighted activeness and amusement, the relationship was not significant. Overall, our results show that people's personality traits and their emotional states are two key elements that predict the tone of their preferred video ads.

Videos are essential for successful crowdfunding campaigns. However, without knowledge of the underlying persuasion factors, novice entrepreneurs may find it difficult to optimize their videos for success. This paper presents VidLyz, a novel assistive tool that allows users to explore the implications of audience-engagement persuasion factors in the context of campaign videos through contrasting examples, crowd-sourced subjective ratings, and feedback on engagement factors. VidLyz promotes active thinking about the impact of audience persuasion factors in making an effective campaign video by guiding novice entrepreneurs in planning materials of their videos on their own, with consideration of the product category, the target audience, and audience-product interactions. To evaluate our system, we collected subjective ratings and feedback on persuasion factors for 140 Kickstarter campaign videos from 2100 crowd workers and presented them through our prototype VidLyz tool. A user study with 45 novice users and five previous campaign creators found that our tool was useful for understanding the implication and relative importance of the persuasion factors of the campaign videos. The interactive and active thinking elements of VidLyz promoted novice users to make coherent and persuasive pre-production plan (using storyboards) for their proposed campaign videos. A follow-up user study showed that these storyboards had a higher likelihood of getting funded by crowd workers than those with less persuasive pre-production plan. We concluded with design implications to better support novice entrepreneurs.

Donation-based crowdfunding platforms have an increasing number of campaigns on socially stigmatized topics. These platforms' widespread online reachability and the large flow of monetary donations have the potential to shape individuals' opinions by influencing their perceptions. However, little research has been done to investigate whether these campaigns impact individuals' opinions and how. We conducted an experiment to explore how an attitude-inconsistent campaign on fairness and equality for LGBTIQ people influenced participants' opinion on this topic. Although all the participants changed their perceived opinions after reading the support for the campaigns, participants opposing equality were less inclined to change their attitude than participants supporting equality. To examine this difference further, we conducted another experiment where participants were exposed to both attitude-consistent and attitude-inconsistent campaigns with varying levels of social support. Participants opposing equality showed less sensitivity to the level of social support, and wanted to donate significantly more money to anti-equality campaigns compared to those who supported equality. Results demonstrate the complex role of crowdfunding campaigns in shaping individuals' opinions on stigmatized topics.

Understanding the factors that persuade backers to donate to research projects has become increasingly important with the rising popularity of scientific crowdfunding. Although there are many similarities between enterprise and scientific crowdfunding, some factors differentiate these two forms of crowdfunding. One such factor is the use of endorsements. The endorsement helps backers gain trust based on expert opinions about the competency of the researchers and the usefulness of the projects. We analyzed 810 endorsements from scientific campaigns posted on Experiment.com and derived a taxonomy of topics discussed in the endorsements. A regression analysis revealed that when endorsers explained the skills of the campaign owners, the probability of success of the campaign improved; on the contrary, when endorsers reiterated the goal of the project, the campaign was less likely to succeed. We conclude with design implications formulated from our findings to better support scientific crowdfunding.

To successfully raise money using crowdfunding, it is important for a campaign to communicate ideas or products effectively to the potential backers. One of the lesser explored but powerful components of a crowdfunding campaign is the campaign video. To better understand how videos affect campaign outcomes, we analyzed videos from 210 Kickstarter campaigns across three different project categories. In a mixed-methods study, we asked 3150 Amazon Mechanical Turk (MTurk) workers to evaluate the campaign videos. We found six recurrent factors from a qualitative analysis as well as quantitative analysis. Analysis revealed product related and video related factors that were predictive of the final outcome of campaigns over and above the static project representation features identified in previous studies. Both the qualitative and quantitative analysis showed that videos influenced perception differently for projects in different categories, and the differential perception was important for predicting successes of the projects. For example, in technology campaigns, projects perceived to have a lower level of complexity were more likely to be successful; but in design and fashion campaigns, projects perceived to have a higher level of complexity - which perhaps reflected craftsmanship - were more likely to be successful. We conclude with design implications to better support the video making process.