Browsing by Author "Reda, Khairi"

Now showing 1 - 10 of 17
  • Thumbnail Image
    Item
    Concept-Driven Visual Analytics: an Exploratory Study of Model- and Hypothesis-Based Reasoning with Visualizations
    (Association for Computer Machinery, 2019) Choi, In Kwon; Childers, Taylor; Raveendranath, Nirmal Kumar; Mishra, Swati; Harris, Kyle; Reda, Khairi
    Visualization tools facilitate exploratory data analysis, but fall short at supporting hypothesis-based reasoning. We conducted an exploratory study to investigate how visualizations might support a concept-driven analysis style, where users can optionally share their hypotheses and conceptual models in natural language, and receive customized plots depicting the fit of their models to the data. We report on how participants leveraged these unique affordances for visual analysis. We found that a majority of participants articulated meaningful models and predictions, utilizing them as entry points to sensemaking. We contribute an abstract typology representing the types of models participants held and externalized as data expectations. Our findings suggest ways for rearchitecting visual analytics tools to better support hypothesis- and model-based reasoning, in addition to their traditional role in exploratory analysis. We discuss the design implications and reflect on the potential benefits and challenges involved.
  • Thumbnail Image
    Item
    Concept-Driven Visual Analytics: an Exploratory Study of Model- and Hypothesis-Based Reasoning with Visualizations
    (ACM, 2019-05) Choi, In Kwon; Childers, Taylor; Raveendranath, Nirmal Kumar; Mishra, Swati; Harris, Kyle; Reda, Khairi; Human-Centered Computing, School of Informatics and Computing
    Visualization tools facilitate exploratory data analysis, but fall short at supporting hypothesis-based reasoning. We conducted an exploratory study to investigate how visualizations might support a concept-driven analysis style, where users can optionally share their hypotheses and conceptual models in natural language, and receive customized plots depicting the fit of their models to the data. We report on how participants leveraged these unique affordances for visual analysis. We found that a majority of participants articulated meaningful models and predictions, utilizing them as entry points to sensemaking. We contribute an abstract typology representing the types of models participants held and externalized as data expectations. Our findings suggest ways for rearchitecting visual analytics tools to better support hypothesis- and model-based reasoning, in addition to their traditional role in exploratory analysis. We discuss the design implications and reflect on the potential benefits and challenges involved.
  • Thumbnail Image
    Item
    Data Prophecy: Exploring the Effects of Belief Elicitation in Visual Analytics
    (2021) Koonchanok, Ratanond; Baser, Parul; Sikharam, Abhinav; Raveendranath, Nirmal Kumar; Reda, Khairi; Human-Centered Computing, School of Informatics and Computing
    Interactive visualizations are widely used in exploratory data analysis, but existing systems provide limited support for confirmatory analysis. We introduce PredictMe, a tool for belief-driven visual analysis, enabling users to draw and test their beliefs against data, as an alternative to data-driven exploration. PredictMe combines belief elicitation with traditional visualization interactions to support mixed analysis styles. In a comparative study, we investigated how these affordances impact participants' cognition. Results show that PredictMe prompts participants to incorporate their working knowledge more frequently in queries. Participants were more likely to attend to discrepancies between their mental models and the data. However, those same participants were also less likely to engage in interactions associated with exploration, and ultimately inspected fewer visualizations and made fewer discoveries. The results suggest that belief elicitation may moderate exploratory behaviors, instead nudging users to be more deliberate in their analysis. We discuss the implications for visualization design.
  • Thumbnail Image
    Item
    Development of a Real-Time Dashboard for Overdose Touchpoints: User-Centered Design Approach
    (JMIR, 2024-06-11) Salvi, Amey; Gillenwater, Logan A.; Cockrum, Brandon P.; Wiehe, Sarah E.; Christian, Kaitlyn; Cayton, John; Bailey, Timothy; Schwartz, Katherine; Dir, Allyson L.; Ray, Bradley; Aalsma, Matthew C.; Reda, Khairi; Pediatrics, School of Medicine
    Background: Overdose Fatality Review (OFR) is an important public health tool for shaping overdose prevention strategies in communities. However, OFR teams review only a few cases at a time, which typically represent a small fraction of the total fatalities in their jurisdiction. Such limited review could result in a partial understanding of local overdose patterns, leading to policy recommendations that do not fully address the broader community needs. Objective: This study explored the potential to enhance conventional OFRs with a data dashboard, incorporating visualizations of touchpoints-events that precede overdoses-to highlight prevention opportunities. Methods: We conducted 2 focus groups and a survey of OFR experts to characterize their information needs and design a real-time dashboard that tracks and measures decedents' past interactions with services in Indiana. Experts (N=27) were engaged, yielding insights on essential data features to incorporate and providing feedback to guide the development of visualizations. Results: The findings highlighted the importance of showing decedents' interactions with health services (emergency medical services) and the justice system (incarcerations). Emphasis was also placed on maintaining decedent anonymity, particularly in small communities, and the need for training OFR members in data interpretation. The developed dashboard summarizes key touchpoint metrics, including prevalence, interaction frequency, and time intervals between touchpoints and overdoses, with data viewable at the county and state levels. In an initial evaluation, the dashboard was well received for its comprehensive data coverage and its potential for enhancing OFR recommendations and case selection. Conclusions: The Indiana touchpoints dashboard is the first to display real-time visualizations that link administrative and overdose mortality data across the state. This resource equips local health officials and OFRs with timely, quantitative, and spatiotemporal insights into overdose risk factors in their communities, facilitating data-driven interventions and policy changes. However, fully integrating the dashboard into OFR practices will likely require training teams in data interpretation and decision-making.
  • Thumbnail Image
    Item
    Dynamic Glyphs: Appropriating Causality Perception in Multivariate Visual Analysis
    (2019) Reda, Khairi; Potts, Caleb; Childers, Taylor; Human-Centered Computing, School of Informatics and Computing
    We investigate how to co-opt the perception of causality to aid the analysis of multivariate data. We propose Dynamic Glyphs (DyGs), an animated extension to traditional glyphs. DyGs encode data relations through seemingly physical interactions between glyph parts. We hypothesize that this representation gives rise to impressions of causality, enabling observers to reason intuitively about complex, multivariate dynamics. In a crowdsourced experiment, participants' accuracy with DyGs exceeded or was comparable to non-animated alternatives. Moreover, participants showed a propensity to infer higher-dimensional relations with DyGs. Our findings suggest that visual causality can be an effective 'channel' for communicating complex data relations that are otherwise difficult to think about. We discuss the implications and highlight future research opportunities.
  • Thumbnail Image
    Item
    Early development of local data dashboards to depict the substance use care cascade for youth involved in the legal system: qualitative findings from end users
    (Springer Nature, 2024-05-30) Dir, Allyson L.; O’Reilly, Lauren; Pederson, Casey; Schwartz, Katherine; Brown, Steven A.; Reda, Khairi; Gillenwater, Logan; Gharbi, Sami; Wiehe, Sarah E.; Adams, Zachary W.; Hulvershorn, Leslie A.; Zapolski, Tamika C. B.; Boustani, Malaz; Aalsma, Matthew C.; Psychiatry, School of Medicine
    Introduction: Rates of substance use are high among youth involved in the legal system (YILS); however, YILS are less likely to initiate and complete substance use treatment compared to their non legally-involved peers. There are multiple steps involved in connecting youth to needed services, from screening and referral within the juvenile legal system to treatment initiation and completion within the behavioral health system. Understanding potential gaps in the care continuum requires data and decision-making from these two systems. The current study reports on the development of data dashboards that integrate these systems' data to help guide decisions to improve substance use screening and treatment for YILS, focusing on end-user feedback regarding dashboard utility. Methods: Three focus groups were conducted with n = 21 end-users from juvenile legal systems and community mental health centers in front-line positions and in decision-making roles across 8 counties to gather feedback on an early version of the data dashboards; dashboards were then modified based on feedback. Results: Qualitative analysis revealed topics related to (1) important aesthetic features of the dashboard, (2) user features such as filtering options and benchmarking to compare local data with other counties, and (3) the centrality of consistent terminology for data dashboard elements. Results also revealed the use of dashboards to facilitate collaboration between legal and behavioral health systems. Conclusions: Feedback from end-users highlight important design elements and dashboard utility as well as the challenges of working with cross-system and cross-jurisdiction data.
  • Thumbnail Image
    Item
    End-User Needs of Fragmented Databases in Higher Education Data Analysis and Decision Making
    (2019-05) Briggs, Amanda; Cafaro, Francesco; Dombrowski, Lynn; Reda, Khairi
    In higher education, a wealth of data is available to advisors, recruiters, marketers, and program directors. However, data sources can be accessed in a variety of ways and often do not seem to represent the same data set, presenting users with the confounding notion that data sources are in conflict with one another. As users are identifying new ways of accessing and analyzing this data, they are modifying existing work practices and sometimes creating their own databases. To understand how users are navigating these databases, the researchers employed a mixed methods research design including a survey and interview to understand the needs to end users who are accessing these seemingly fragmented databases. The study resulted in a three overarching categories – access, understandability, and use – that affect work practices for end users. The researchers used these themes to develop a set of broadly applicable design recommendations as well as six sets of sketches for implementation – development of a data gateway, training, collaboration, tracking, definitions and roadblocks, and time management.
  • Thumbnail Image
    Item
    Evaluating Gradient Perception in Color-Coded Scalar Fields
    (IEEE, 2019-10) Reda, Khairi; Papka, Michael E.; Human-Centered Computing, School of Informatics and Computing
    Color mapping is a commonly used technique for visualizing scalar fields. While there exists advice for choosing effective colormaps, it is unclear if current guidelines apply equally across task types. We study the perception of gradients and evaluate the effectiveness of three colormaps at depicting gradient magnitudes. In a crowd-sourced experiment, we determine the just-noticeable differences (JNDs) at which participants can reliably compare and judge variations in gradient between two scalar fields. We find that participants exhibited lower JNDs with a diverging (cool-warm) or a spectral (rainbow) scheme, as compared with a monotonic-luminance colormap (viridis). The results support a hypothesis that apparent discontinuities in the color ramp may help viewers discern subtle structural differences in gradient. We discuss these findings and highlight future research directions for colormap evaluation.
  • Thumbnail Image
    Item
    Graphical Perception of Continuous Quantitative Maps: the Effects of Spatial Frequency and Colormap Design
    (ACM, 2018) Reda, Khairi; Nalawade, Pratik; Ansah-Koi, Kate
    Continuous 'pseudocolor' maps visualize how a quantitative attribute varies smoothly over space. These maps are widely used by experts and lay citizens alike for communicating scientific and geographical data. A critical challenge for designers of these maps is selecting a color scheme that is both effective and aesthetically pleasing. Although there exist empirically grounded guidelines for color choice in segmented maps (e.g., choropleths), continuous maps are significantly understudied, and their color-coding guidelines are largely based on expert opinion and design heuristics--many of these guidelines have yet to be verified experimentally. We conducted a series of crowdsourced experiments to investigate how the perception of continuous maps is affected by colormap characteristics and spatial frequency (a measure of data complexity). We find that spatial frequency significantly impacts the effectiveness of color encodes, but the precise effect is task-dependent. While rainbow schemes afforded the highest accuracy in quantity estimation irrespective of spatial complexity, divergent colormaps significantly outperformed other schemes in tasks requiring the perception of high-frequency patterns. We interpret these results in relation to current practices and devise new and more granular guidelines for color mapping in continuous maps.
  • Thumbnail Image
    Item
    Into the Black Box: Designing for Transparency in Artificial Intelligence
    (2019-11) Vorm, Eric Stephen; Miller, Andrew; Bolchini, Davide; Reda, Khairi; Fedorikhin, Sasha
    The rapid infusion of artificial intelligence into everyday technologies means that consumers are likely to interact with intelligent systems that provide suggestions and recommendations on a daily basis in the very near future. While these technologies promise much, current issues in low transparency create high potential to confuse end-users, limiting the market viability of these technologies. While efforts are underway to make machine learning models more transparent, HCI currently lacks an understanding of how these model-generated explanations should best translate into the practicalities of system design. To address this gap, my research took a pragmatic approach to improving system transparency for end-users. Through a series of three studies, I investigated the need and value of transparency to end-users, and explored methods to improve system designs to accomplish greater transparency in intelligent systems offering recommendations. My research resulted in a summarized taxonomy that outlines a variety of motivations for why users ask questions of intelligent systems; useful for considering the type and category of information users might appreciate when interacting with AI-based recommendations. I also developed a categorization of explanation types, known as explanation vectors, that is organized into groups that correspond to user knowledge goals. Explanation vectors provide system designers options for delivering explanations of system processes beyond those of basic explainability. I developed a detailed user typology, which is a four-factor categorization of the predominant attitudes and opinion schemes of everyday users interacting with AI-based recommendations; useful to understand the range of user sentiment towards AI-based recommender features, and possibly useful for tailoring interface design by user type. Lastly, I developed and tested an evaluation method known as the System Transparency Evaluation Method (STEv), which allows for real-world systems and prototypes to be evaluated and improved through a low-cost query method. Results from this dissertation offer concrete direction to interaction designers as to how these results might manifest in the design of interfaces that are more transparent to end users. These studies provide a framework and methodology that is complementary to existing HCI evaluation methods, and lay the groundwork upon which other research into improving system transparency might build.