Illusion of Causality in Visualized Data

How Aligned are Human Chart Takeaways and LLM Predictions? A Case Study on Bar Charts with Varying Layouts

Large Language Models (LLMs) have been adopted for a variety of visualizations tasks, but how far are we from perceptually aware LLMs that can predict human takeaways? Graphical perception literature has shown that human chart takeaways are sensitive to visualization design choices, such as spatial layouts. In this work, we examine the extent to which LLMs exhibit such sensitivity when generating takeaways, using bar charts with varying spatial layouts as a case study. We conducted three experiments and tested four common bar chart layouts: vertically juxtaposed, horizontally juxtaposed, overlaid, and stacked. In Experiment 1, we identified the optimal configurations to generate meaningful chart takeaways by testing four LLMs, two temperature settings, nine chart specifications, and two prompting strategies. We found that even state-of-the-art LLMs struggled to generate semantically diverse and factually accurate takeaways. In Experiment 2, we used the optimal configurations to generate 30 chart takeaways each for eight visualizations across four layouts and two datasets in both zero-shot and one-shot settings. Compared to human takeaways, we found that the takeaways LLMs generated often did not match the types of comparisons made by humans. In Experiment 3, we examined the effect of chart context and data on LLM takeaways. We found that LLMs, unlike humans, exhibited variation in takeaway comparison types for different bar charts using the same bar layout. Overall, our case study evaluates the ability of LLMs to emulate human interpretations of data and points to challenges and opportunities in using LLMs to predict human chart takeaways.

Beyond Correlation: Incorporating Counterfactual Guidance to Better Support Exploratory Visual Analysis

Providing effective guidance for users has long been an important and challenging task for efficient exploratory visual analytics, especially when selecting variables for visualization in high-dimensional datasets. Correlation is the most widely applied metric for guidance in statistical and analytical tools, however a reliance on correlation may lead users towards false positives when interpreting causal relations in the data. In this work, inspired by prior insights on the benefits of counterfactual visualization in supporting visual causal inference, we propose a novel, simple, and efficient counterfactual guidance method to enhance causal inference performance in guided exploratory analytics based on insights and concerns gathered from expert interviews. Our technique aims to capitalize on the benefits of counterfactual approaches while reducing their complexity for users. We integrated counterfactual guidance into an exploratory visual analytics system, and using a synthetically generated ground-truth causal dataset, conducted a comparative user study and evaluated to what extent counterfactual guidance can help lead users to more precise visual causal inferences. The results suggest that counterfactual guidance improved visual causal inference performance, and also led to different exploratory behaviors compared to correlation-based guidance. Based on these findings, we offer future directions and challenges for incorporating counterfactual guidance to better support exploratory visual analytics.

Causal Priors and Their Influence on Judgements of Causality in Visualized Data

"Correlation does not imply causation" is a famous mantra in statistical and visual analysis. However, consumers of visualizations often draw causal conclusions when only correlations between variables are shown. In this paper, we investigate factors that contribute to causal relationships users perceive in visualizations. We collected a corpus of concept pairs from variables in widely used datasets and created visualizations that depict varying correlative associations using three typical statistical chart types. We conducted two MTurk studies on (1) preconceived notions on causal relations without charts, and (2) perceived causal relations with charts, for each concept pair. Our results indicate that people make assumptions about causal relationships between pairs of concepts even without seeing any visualized data. Moreover, our results suggest that these assumptions constitute causal priors that, in combination with visualized association, impact how data visualizations are interpreted. The results also suggest that causal priors may lead to over- or under-estimation in perceived causal relations in different circumstances, and that those priors can also impact users' confidence in their causal assessments. In addition, our results align with prior work, indicating that chart type may also affect causal inference. Using data from the studies, we develop a model to capture the interaction between causal priors and visualized associations as they combine to impact a user's perceived causal relations. In addition to reporting the study results and analyses, we provide an open dataset of causal priors for 56 specific concept pairs that can serve as a potential benchmark for future studies. We also suggest remaining challenges and heuristic-based guidelines to help designers improve visualization design choices to better support visual causal inference.

Decoupling Judgment and Decision Making: A Tale of Two Tails

Is it true that if citizens understand hurricane probabilities, they will make more rational decisions for evacuation? Finding answers to such questions is not straightforward in the literature because the terms "judgment" and "decision making" are often used interchangeably. This terminology conflation leads to a lack of clarity on whether people make suboptimal decisions because of inaccurate judgments of information conveyed in visualizations or because they use alternative yet currently unknown heuristics. To decouple judgment from decision making, we review relevant concepts from the literature and present two preregistered experiments (N = 601) to investigate if the task (judgment versus decision making), the scenario (sports versus humanitarian), and the visualization (quantile dotplots, density plots, probability bars) affect accuracy. While experiment 1 was inconclusive, we found evidence for a difference in experiment 2. Contrary to our expectations and previous research, which found decisions less accurate than their direct-equivalent judgments, our results pointed in the opposite direction. Our findings further revealed that decisions were less vulnerable to status-quo bias, suggesting decision makers may disfavor responses associated with inaction. We also found that both scenario and visualization types can influence people's judgments and decisions. Although effect sizes are not large and results should be interpreted carefully, we conclude that judgments cannot be safely used as proxy tasks for decision making, and discuss implications for visualization research and beyond. Materials and preregistrations are available at https://osf.io/ufzp5/?view_only=adc0f78a23804c31bf7fdd9385cb264f.

MediVizor: Visual Mediation Analysis of Nominal Variables

Mediation analysis is crucial for diagnosing indirect causal relations in many scientific fields. However, mediation analysis of nominal variables requires examining and comparing multiple total effects and their corresponding direct/indirect causal effects derived from mediation models. This process is tedious and challenging to achieve with classical analysis tools such as Excel tables. In this study, we worked closely with experts from two scientific domains to design MediVizor, a visualization system that enables experts to conduct visual mediation analysis of nominal variables. The visualization design allows users to browse and compare multiple total effects together with the direct/indirect effects that compose them. The design also allows users to examine to what extent the positive and negative direct/indirect effects contribute to and reduce the total effects, respectively. We conducted two case studies separately with the experts from the two domains, sports and communication science, and a user study with common users to evaluate the system and design. The positive feedback from experts and common users demonstrates the effectiveness and generalizability of the system.

Same Data, Diverging Perspectives: The Power of Visualizations to Elicit Competing Interpretations

People routinely rely on data to make decisions, but the process can be riddled with biases. We show that patterns in data might be noticed first or more strongly, depending on how the data is visually represented or what the viewer finds salient. We also demonstrate that viewer interpretation of data is similar to that of 'ambiguous figures' such that two people looking at the same data can come to different decisions. In our studies, participants read visualizations depicting competitions between two entities, where one has a historical lead (A) but the other has been gaining momentum (B) and predicted a winner, across two chart types and three annotation approaches. They either saw the historical lead as salient and predicted that A would win, or saw the increasing momentum as salient and predicted B to win. These results suggest that decisions can be influenced by both how data are presented and what patterns people find visually salient.

Causality-Based Visual Analysis of Questionnaire Responses

As the final stage of questionnaire analysis, causal reasoning is the key to turning responses into valuable insights and actionable items for decision-makers. During the questionnaire analysis, classical statistical methods (e.g., Differences-in-Differences) have been widely exploited to evaluate causality between questions. However, due to the huge search space and complex causal structure in data, causal reasoning is still extremely challenging and time-consuming, and often conducted in a trial-and-error manner. On the other hand, existing visual methods of causal reasoning face the challenge of bringing scalability and expert knowledge together and can hardly be used in the questionnaire scenario. In this work, we present a systematic solution to help analysts effectively and efficiently explore questionnaire data and derive causality. Based on the association mining algorithm, we dig question combinations with potential inner causality and help analysts interactively explore the causal sub-graph of each question combination. Furthermore, leveraging the requirements collected from the experts, we built a visualization tool and conducted a comparative study with the state-of-the-art system to show the usability and efficiency of our system.

Polarizing Political Polls: How Visualization Design Choices Can Shape Public Opinion and Increase Political Polarization

While we typically focus on data visualization as a tool for facilitating cognitive tasks (e.g. learning facts, making decisions), we know relatively little about their second-order impacts on our opinions, attitudes, and values. For example, could design or framing choices interact with viewers' social cognitive biases in ways that promote political polarization? When reporting on U.S. attitudes toward public policies, it is popular to highlight the gap between Democrats and Republicans (e.g. with blue vs red connected dot plots). But these charts may encourage social-normative conformity, influencing viewers' attitudes to match the divided opinions shown in the visualization. We conducted three experiments examining visualization framing in the context of social conformity and polarization. Crowdworkers viewed charts showing simulated polling results for public policy proposals. We varied framing (aggregating data as non-partisan "All US Adults," or partisan "Democrat" / "Republican") and the visualized groups' support levels. Participants then reported their own support for each policy. We found that participants' attitudes biased significantly toward the group attitudes shown in the stimuli and this can increase inter-party attitude divergence. These results demonstrate that data visualizations can induce social conformity and accelerate political polarization. Choosing to visualize partisan divisions can divide us further.

Using Counterfactuals to Improve Causal Inferences From Visualizations

Traditional approaches to data visualization have often focused on comparing different subsets of data, and this is reflected in the many techniques developed and evaluated over the years for visual comparison. Similarly, common workflows for exploratory visualization are built upon the idea of users interactively applying various filter and grouping mechanisms in search of new insights. This paradigm has proven effective at helping users identify correlations between variables that can inform thinking and decision-making. However, recent studies show that consumers of visualizations often draw causal conclusions even when not supported by the data. Motivated by these observations, this article highlights recent advances from a growing community of researchers exploring methods that aim to directly support visual causal inference. However, many of these approaches have their own limitations, which limit their use in many real-world scenarios. This article, therefore, also outlines a set of key open challenges and corresponding priorities for new research to advance the state of the art in visual causal inference.

My Model is Unfair, Do People Even Care? Visual Design Affects Trust and Perceived Bias in Machine Learning

Machine learning technology has become ubiquitous, but, unfortunately, often exhibits bias. As a consequence, disparate stakeholders need to interact with and make informed decisions about using machine learning models in everyday systems. Visualization technology can support stakeholders in understanding and evaluating trade-offs between, for example, accuracy and fairness of models. This paper aims to empirically answer "Can visualization design choices affect a stakeholder's perception of model bias, trust in a model, and willingness to adopt a model?" Through a series of controlled, crowd-sourced experiments with more than 1,500 participants, we identify a set of strategies people follow in deciding which models to trust. Our results show that men and women prioritize fairness and performance differently and that visual design choices significantly affect that prioritization. For example, women trust fairer models more often than men do, participants value fairness more when it is explained using text than as a bar chart, and being explicitly told a model is biased has a bigger impact than showing past biased performance. We test the generalizability of our results by comparing the effect of multiple textual and visual design choices and offer potential explanations of the cognitive mechanisms behind the difference in fairness perception and trust. Our research guides design considerations to support future work developing visualization systems for machine learning.

DASS Good: Explainable Data Mining of Spatial Cohort Data

Developing applicable clinical machine learning models is a difficult task when the data includes spatial information, for example, radiation dose distributions across adjacent organs at risk. We describe the co-design of a modeling system, DASS, to support the hybrid human-machine development and validation of predictive models for estimating long-term toxicities related to radiotherapy doses in head and neck cancer patients. Developed in collaboration with domain experts in oncology and data mining, DASS incorporates human-in-the-loop visual steering, spatial data, and explainable AI to augment domain knowledge with automatic data mining. We demonstrate DASS with the development of two practical clinical stratification models and report feedback from domain experts. Finally, we describe the design lessons learned from this collaborative experience.

How Do Viewers Synthesize Conflicting Information from Data Visualizations?

Scientific knowledge develops through cumulative discoveries that build on, contradict, contextualize, or correct prior findings. Scientists and journalists often communicate these incremental findings to lay people through visualizations and text (e.g., the positive and negative effects of caffeine intake). Consequently, readers need to integrate diverse and contrasting evidence from multiple sources to form opinions or make decisions. However, the underlying mechanism for synthesizing information from multiple visualizations remains under-explored. To address this knowledge gap, we conducted a series of four experiments ( N=1166) in which participants synthesized empirical evidence from a pair of line charts presented sequentially. In Experiment 1, we administered a baseline condition with charts depicting no specific context where participants held no strong belief. To test for the generalizability, we introduced real-world scenarios to our visualizations in Experiment 2 and added accompanying text descriptions similar to online news articles or blog posts in Experiment 3. In all three experiments, we varied the relative direction and magnitude of line slopes within the chart pairs. We found that participants tended to weigh the positive slope more when the two charts depicted relationships in the opposite direction (e.g., one positive slope and one negative slope). Participants tended to weigh the less steep slope more when the two charts depicted relationships in the same direction (e.g., both positive). Through these experiments, we characterize participants' synthesis behaviors depending on the relationship between the information they viewed, contribute to theories describing underlying cognitive mechanisms in information synthesis, and describe design implications for data storytelling.

Comparison Conundrum and the Chamber of Visualizations: An Exploration of How Language Influences Visual Design

The language for expressing comparisons is often complex and nuanced, making supporting natural language-based visual comparison a non-trivial task. To better understand how people reason about comparisons in natural language, we explore a design space of utterances for comparing data entities. We identified different parameters of comparison utterances that indicate what is being compared (i.e., data variables and attributes) as well as how these parameters are specified (i.e., explicitly or implicitly). We conducted a user study with sixteen data visualization experts and non-experts to investigate how they designed visualizations for comparisons in our design space. Based on the rich set of visualization techniques observed, we extracted key design features from the visualizations and synthesized them into a subset of sixteen representative visualization designs. We then conducted a follow-up study to validate user preferences for the sixteen representative visualizations corresponding to utterances in our design space. Findings from these studies suggest guidelines and future directions for designing natural language interfaces and recommendation tools to better support natural language comparisons in visual analytics.

Seeing What You Believe or Believing What You See? Belief Biases Correlation Estimation

When an analyst or scientist has a belief about how the world works, their thinking can be biased in favor of that belief. Therefore, one bedrock principle of science is to minimize that bias by testing the predictions of one's belief against objective data. But interpreting visualized data is a complex perceptual and cognitive process. Through two crowdsourced experiments, we demonstrate that supposedly objective assessments of the strength of a correlational relationship can be influenced by how strongly a viewer believes in the existence of that relationship. Participants viewed scatterplots depicting a relationship between meaningful variable pairs (e.g., number of environmental regulations and air quality) and estimated their correlations. They also estimated the correlation of the same scatterplots labeled instead with generic 'X' and 'Y' axes. In a separate section, they also reported how strongly they believed there to be a correlation between the meaningful variable pairs. Participants estimated correlations more accurately when they viewed scatterplots labeled with generic axes compared to scatterplots labeled with meaningful variable pairs. Furthermore, when viewers believed that two variables should have a strong relationship, they overestimated correlations between those variables by an r-value of about 0.1. When they believed that the variables should be unrelated, they underestimated the correlations by an r-value of about 0.1. While data visualizations are typically thought to present objective truths to the viewer, these results suggest that existing personal beliefs can bias even objective statistical values people extract from data.

Dispersion vs Disparity: Hiding Variability Can Encourage Stereotyping When Visualizing Social Outcomes

Visualization research often focuses on perceptual accuracy or helping readers interpret key messages. However, we know very little about how chart designs might influence readers' perceptions of the people behind the data. Specifically, could designs interact with readers' social cognitive biases in ways that perpetuate harmful stereotypes? For example, when analyzing social inequality, bar charts are a popular choice to present outcome disparities between race, gender, or other groups. But bar charts may encourage deficit thinking, the perception that outcome disparities are caused by groups' personal strengths or deficiencies, rather than external factors. These faulty personal attributions can then reinforce stereotypes about the groups being visualized. We conducted four experiments examining design choices that influence attribution biases (and therefore deficit thinking). Crowdworkers viewed visualizations depicting social outcomes that either mask variability in data, such as bar charts or dot plots, or emphasize variability in data, such as jitter plots or prediction intervals. They reported their agreement with both personal and external explanations for the visualized disparities. Overall, when participants saw visualizations that hide within-group variability, they agreed more with personal explanations. When they saw visualizations that emphasize within-group variability, they agreed less with personal explanations. These results demonstrate that data visualizations about social inequity can be misinterpreted in harmful ways and lead to stereotyping. Design choices can influence these biases: Hiding variability tends to increase stereotyping while emphasizing variability reduces it.

Causal theory error in college students' understanding of science studies

When reasoning about science studies, people often make causal theory errors by inferring or accepting a causal claim based on correlational evidence. While humans naturally think in terms of causal relationships, reasoning about science findings requires understanding how evidence supports—or fails to support—a causal claim. This study investigated college students’ thinking about causal claims presented in brief media reports describing behavioral science findings. How do science students reason about causal claims from correlational evidence? And can their reasoning be improved through instruction clarifying the nature of causal theory error? We examined these questions through a series of written reasoning exercises given to advanced college students over three weeks within a psychology methods course. In a pretest session, students critiqued study quality and support for a causal claim from a brief media report  suggesting an association between two variables. Then, they created diagrams depicting possible alternative causal theories. At the beginning of the second session, an instructional intervention introduced students to an extended example of a causal theory error through guided questions about possible alternative causes. Then, they completed the same two tasks with new science reports immediately and again 1 week later. The results show students’ reasoning included fewer causal theory errors after the intervention, and this improvement was maintained a week later. Our findings suggest that interventions aimed at addressing reasoning about causal claims in correlational studies are needed even for advanced science students, and that training on considering alternative causal theories may be successful in reducing casual theory error.

Compass: Towards Better Causal Analysis of Urban Time Series

The spatial time series generated by city sensors allow us to observe urban phenomena like environmental pollution and traffic congestion at an unprecedented scale. However, recovering causal relations from these observations to explain the sources of urban phenomena remains a challenging task because these causal relations tend to be time-varying and demand proper time series partitioning for effective analyses. The prior approaches extract one causal graph given long-time observations, which cannot be directly applied to capturing, interpreting, and validating dynamic urban causality. This paper presents Compass, a novel visual analytics approach for in-depth analyses of the dynamic causality in urban time series. To develop Compass, we identify and address three challenges: detecting urban causality, interpreting dynamic causal relations, and unveiling suspicious causal relations. First, multiple causal graphs over time among urban time series are obtained with a causal detection framework extended from the Granger causality test. Then, a dynamic causal graph visualization is designed to reveal the time-varying causal relations across these causal graphs and facilitate the exploration of the graphs along the time. Finally, a tailored multi-dimensional visualization is developed to support the identification of spurious causal relations, thereby improving the reliability of causal analyses. The effectiveness of Compass is evaluated with two case studies conducted on the real-world urban datasets, including the air pollution and traffic speed datasets, and positive feedback was received from domain experts.

Visual Arrangements of Bar Charts Influence Comparisons in Viewer Takeaways

Well-designed data visualizations can lead to more powerful and intuitive processing by a viewer. To help a viewer intuitively compare values to quickly generate key takeaways, visualization designers can manipulate how data values are arranged in a chart to afford particular comparisons. Using simple bar charts as a case study, we empirically tested the comparison affordances of four common arrangements: vertically juxtaposed, horizontally juxtaposed, overlaid, and stacked. We asked participants to type out what patterns they perceived in a chart and we coded their takeaways into types of comparisons. In a second study, we asked data visualization design experts to predict which arrangement they would use to afford each type of comparison and found both alignments and mismatches with our findings. These results provide concrete guidelines for how both human designers and automatic chart recommendation systems can make visualizations that help viewers extract the "right" takeaway.

The Science of Visual Data Communication: What Works

Effectively designed data visualizations allow viewers to use their powerful visual systems to understand patterns in data across science, education, health, and public policy. But ineffectively designed visualizations can cause confusion, misunderstanding, or even distrust-especially among viewers with low graphical literacy. We review research-backed guidelines for creating effective and intuitive visualizations oriented toward communicating data to students, coworkers, and the general public. We describe how the visual system can quickly extract broad statistics from a display, whereas poorly designed displays can lead to misperceptions and illusions. Extracting global statistics is fast, but comparing between subsets of values is slow. Effective graphics avoid taxing working memory, guide attention, and respect familiar conventions. Data visualizations can play a critical role in teaching and communication, provided that designers tailor those visualizations to their audience.

Once Upon A Time In Visualization: Understanding the Use of Textual Narratives for Causality

Causality visualization can help people understand temporal chains of events, such as messages sent in a distributed system, cause and effect in a historical conflict, or the interplay between political actors over time. However, as the scale and complexity of these event sequences grows, even these visualizations can become overwhelming to use. In this paper, we propose the use of textual narratives as a data-driven storytelling method to augment causality visualization. We first propose a design space for how textual narratives can be used to describe causal data. We then present results from a crowdsourced user study where participants were asked to recover causality information from two causality visualizations-causal graphs and Hasse diagrams-with and without an associated textual narrative. Finally, we describe Causeworks, a causality visualization system for understanding how specific interventions influence a causal model. The system incorporates an automatic textual narrative mechanism based on our design space. We validate Causeworks through interviews with experts who used the system for understanding complex events.

A Visual Analytics Approach for Exploratory Causal Analysis: Exploration, Validation, and Applications

Using causal relations to guide decision making has become an essential analytical task across various domains, from marketing and medicine to education and social science. While powerful statistical models have been developed for inferring causal relations from data, domain practitioners still lack effective visual interface for interpreting the causal relations and applying them in their decision-making process. Through interview studies with domain experts, we characterize their current decision-making workflows, challenges, and needs. Through an iterative design process, we developed a visualization tool that allows analysts to explore, validate, and apply causal relations in real-world decision-making scenarios. The tool provides an uncertainty-aware causal graph visualization for presenting a large set of causal relations inferred from high-dimensional data. On top of the causal graph, it supports a set of intuitive user controls for performing what-if analyses and making action plans. We report on two case studies in marketing and student advising to demonstrate that users can effectively explore causal relations and design action plans for reaching their goals.