Symmetry and spatial ability enhance change detection in visuospatial structures.

Science, Technology, Engineering, and Mathematics (STEM) domains require people to recognize and transform complex visuospatial displays that appear to vastly exceed the limits of visuospatial working memory. Here, we consider possible domain-general mechanisms that may explain this advantage: capitalizing on symmetry, a structural regularity that can produce more efficient representations. Participants briefly viewed a structure made up of three-dimensional connected cubes of different colors, which was either asymmetrical or symmetrical. After a short delay, they were asked to detect a change (colors swapping positions) within a rotated second view. In change trials, the second display always had an asymmetrical structure. The presence of symmetry in the initial view improved change detection, and performance also declined with angular disparity of the encoding and test displays. People with higher spatial ability performed better on the change-detection task, but there was no evidence that they were better at leveraging symmetry than low-spatial individuals. The results suggest that leveraging symmetrical structures can help people of all ability levels exceed typical working memory limits by constructing more efficient representations and substituting resource-demanding mental rotation operations with alternative orientation-independent strategies.

Similarity Grouping as Feature-Based Selection.

Across the natural world as well as the artificial worlds of maps, diagrams, and data visualizations, feature similarity (e.g., color and shape) links spatially separate areas into sets. Despite a century of study, it is yet unclear what mechanism underlies this gestalt similarity grouping. One recent proposal is that similarity grouping-for example, seeing a red, vertical, or square group-is just global selection of those features. Although parsimonious, this account makes the counterintuitive prediction that similarity grouping is strictly serial: A green group cannot be constructed at the same time as a red group. We tested this prediction with a novel measure-a grouping illusion within number-estimation tasks that should work only if participants simultaneously construct groups-and found the strongest evidence yet in favor of serial feature-based attention ( Ns = 14, 12, and 12 for Experiment 1, Experiment 2, and Experiment 3, respectively).

Gesture helps learners learn, but not merely by guiding their visual attention.

Teaching a new concept through gestures-hand movements that accompany speech-facilitates learning above-and-beyond instruction through speech alone (e.g., Singer & Goldin-Meadow, ). However, the mechanisms underlying this phenomenon are still under investigation. Here, we use eye tracking to explore one often proposed mechanism-gesture's ability to direct visual attention. Behaviorally, we replicate previous findings: Children perform significantly better on a posttest after learning through Speech+Gesture instruction than through Speech Alone instruction. Using eye tracking measures, we show that children who watch a math lesson with gesture do allocate their visual attention differently from children who watch a math lesson without gesture-they look more to the problem being explained, less to the instructor, and are more likely to synchronize their visual attention with information presented in the instructor's speech (i.e., follow along with speech) than children who watch the no-gesture lesson. The striking finding is that, even though these looking patterns positively predict learning outcomes, the patterns do not mediate the effects of training condition (Speech Alone vs. Speech+Gesture) on posttest success. We find instead a complex relation between gesture and visual attention in which gesture moderates the impact of visual looking patterns on learning-following along with speech predicts learning for children in the Speech+Gesture condition, but not for children in the Speech Alone condition. Gesture's beneficial effects on learning thus come not merely from its ability to guide visual attention, but also from its ability to synchronize with speech and affect what learners glean from that speech.

Topological Relations Between Objects Are Categorically Coded.

How do individuals compare images-for example, two graphs or diagrams-to identify differences between them? We argue that categorical relations between objects play a critical role. These relations divide continuous space into discrete categories, such as "above" and "below," or "containing" and "overlapping," which are remembered and compared more easily than precise metric values. These relations should lead to categorical perception, such that viewers find it easier to notice a change that crosses a category boundary (one object is now above, rather than below, another, or now contains, rather than overlaps with, another) than a change of equal magnitude that does not cross a boundary. We tested the influence of a set of topological categorical relations from the cognitive-modeling literature. In a visual same/different comparison task, viewers more accurately noticed changes that crossed relational category boundaries, compared with changes that did not cross these boundaries. The results highlight the potential of systematic exploration of the boundaries of between-object relational categories.

Four types of ensemble coding in data visualizations.

Ensemble coding supports rapid extraction of visual statistics about distributed visual information. Researchers typically study this ability with the goal of drawing conclusions about how such coding extracts information from natural scenes. Here we argue that a second domain can serve as another strong inspiration for understanding ensemble coding: graphs, maps, and other visual presentations of data. Data visualizations allow observers to leverage their ability to perform visual ensemble statistics on distributions of spatial or featural visual information to estimate actual statistics on data. We survey the types of visual statistical tasks that occur within data visualizations across everyday examples, such as scatterplots, and more specialized images, such as weather maps or depictions of patterns in text. We divide these tasks into four categories: identification of sets of values, summarization across those values, segmentation of collections, and estimation of structure. We point to unanswered questions for each category and give examples of such cross-pollination in the current literature. Increased collaboration between the data visualization and perceptual psychology research communities can inspire new solutions to challenges in visualization while simultaneously exposing unsolved problems in perception research.

Capacity for Visual Features in Mental Rotation.

Although mental rotation is a core component of scientific reasoning, little is known about its underlying mechanisms. For instance, how much visual information can someone rotate at once? We asked participants to rotate a simple multipart shape, requiring them to maintain attachments between features and moving parts. The capacity of this aspect of mental rotation was strikingly low: Only one feature could remain attached to one part. Behavioral and eye-tracking data showed that this single feature remained "glued" via a singular focus of attention, typically on the object's top. We argue that the architecture of the human visual system is not suited for keeping multiple features attached to multiple parts during mental rotation. Such measurement of capacity limits may prove to be a critical step in dissecting the suite of visuospatial tools involved in mental rotation, leading to insights for improvement of pedagogy in science-education contexts.

Anxious mood narrows attention in feature space.

Spatial attention can operate like a spotlight whose scope can vary depending on task demands. Emotional states contribute to the spatial extent of attentional selection, with the spotlight focused more narrowly during anxious moods and more broadly during happy moods. In addition to visual space, attention can also operate over features, and we show here that mood states may also influence attentional scope in feature space. After anxious or happy mood inductions, participants focused their attention to identify a central target while ignoring flanking items. Flankers were sometimes coloured differently than targets, so focusing attention on target colour should lead to relatively less interference. Compared to happy and neutral moods, when anxious, participants showed reduced interference when colour isolated targets from flankers, but showed more interference when flankers and targets were the same colour. This pattern reveals that the anxious mood caused these individuals to attend to the irrelevant feature in both cases, regardless of its benefit or detriment. In contrast, participants showed no effect of colour on interference when happy, suggesting that positive mood did not influence attention in feature space. These mood effects on feature-based attention provide a theoretical bridge between previous findings concerning spatial and conceptual attention.

Reading a graph is like reading a paragraph.

Vision provides rapid processing for some tasks, but encounters strong constraints from others. Although many tasks encounter a capacity limit of processing four visual objects at once, some evidence suggests far lower limits for processing relationships among objects. What is our capacity limit for relational processing? If it is indeed limited, then people may miss important relationships between data values in a graph. To test this question, we asked people to explore graphs of trivially simple 2 × 2 data sets and found that half of the viewers missed surprising and improbable relationships (e.g., a child's height decreasing over time). These relationships were spotted easily in a control condition, which implicitly directed viewers to prioritize inspecting the key relationships. Thus, a severe limit on relational processing, combined with a cascade of other capacity-limited operations (e.g., linking values to semantic content), makes understanding a graph more like slowly reading a paragraph then immediately recognizing an image. These results also highlight the practical importance of "data storytelling" techniques, where communicators design graphs that help their audience prioritize the most important relationships in data. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

The Arrangement of Marks Impacts Afforded Messages: Ordering, Partitioning, Spacing, and Coloring in Bar Charts.

Data visualizations present a massive number of potential messages to an observer. One might notice that one group's average is larger than another's, or that a difference in values is smaller than a difference between two others, or any of a combinatorial explosion of other possibilities. The message that a viewer tends to notice - the message that a visualization 'affords' - is strongly affected by how values are arranged in a chart, e.g., how the values are colored or positioned. Although understanding the mapping between a chart's arrangement and what viewers tend to notice is critical for creating guidelines and recommendation systems, current empirical work is insufficient to lay out clear rules. We present a set of empirical evaluations of how different messages-including ranking, grouping, and part-to-whole relationships-are afforded by variations in ordering, partitioning, spacing, and coloring of values, within the ubiquitous case study of bar graphs. In doing so, we introduce a quantitative method that is easily scalable, reviewable, and replicable, laying groundwork for further investigation of the effects of arrangement on message affordances across other visualizations and tasks. Pre-registration and all supplemental materials are available at https://osf.io/np3q7 and https://osf.io/bvy95, respectively.

Language systematizes attention: How relational language enhances relational representation by guiding attention.

Language can affect cognition, but through what mechanism? Substantial past research has focused on how labeling can elicit categorical representation during online processing. We focus here on a particularly powerful type of language-relational language-and show that relational language can enhance relational representation in children through an embodied attention mechanism. Four-year-old children were given a color-location conjunction task, in which they were asked to encode a two-color square, split either vertically or horizontally (e.g., red on the left, blue on the right), and later recall the same configuration from its mirror reflection. During the encoding phase, children in the experimental condition heard relational language (e.g., "Red is on the left of blue"), while those in the control condition heard generic non-relational language (e.g., "Look at this one, look at it closely"). At recall, children in the experimental condition were more successful at choosing the correct relational representation between the two colors compared to the control group. Moreover, they exhibited different attention patterns as predicted by the attention shift account of relational representation (Franconeri et al., 2012). To test the sustained effect of language and the role of attention, during the second half of the study, the experimental condition was given generic non-relational language. There was a sustained advantage in the experimental condition for both behavioral accuracies and signature attention patterns. Overall, our findings suggest that relational language enhances relational representation by guiding learners' attention, and this facilitative effect persists over time even in the absence of language. Implications for the mechanism of how relational language can enhance the learning of relational systems (e.g., mathematics, spatial cognition) by guiding attention will be discussed.

Average Estimates in Line Graphs Are Biased Toward Areas of Higher Variability.

We investigate variability overweighting, a previously undocumented bias in line graphs, where estimates of average value are biased toward areas of higher variability in that line. We found this effect across two preregistered experiments with 140 and 420 participants. These experiments also show that the bias is reduced when using a dot encoding of the same series. We can model the bias with the average of the data series and the average of the points drawn along the line. This bias might arise because higher variability leads to stronger weighting in the average calculation, either due to the longer line segments (even though those segments contain the same number of data values) or line segments with higher variability being otherwise more visually salient. Understanding and predicting this bias is important for visualization design guidelines, recommendation systems, and tool builders, as the bias can adversely affect estimates of averages and trends.

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.

Swaying the Public? Impacts of Election Forecast Visualizations on Emotion, Trust, and Intention in the 2022 U.S. Midterms.

We conducted a longitudinal study during the 2022 U.S. midterm elections, investigating the real-world impacts of uncertainty visualizations. Using our forecast model of the governor elections in 33 states, we created a website and deployed four uncertainty visualizations for the election forecasts: single quantile dotplot (1-Dotplot), dual quantile dotplots (2-Dotplot), dual histogram intervals (2-Interval), and Plinko quantile dotplot (Plinko), an animated design with a physical and probabilistic analogy. Our online experiment ran from Oct. 18, 2022, to Nov. 23, 2022, involving 1,327 participants from 15 states. We use Bayesian multilevel modeling and post-stratification to produce demographically-representative estimates of people's emotions, trust in forecasts, and political participation intention. We find that election forecast visualizations can heighten emotions, increase trust, and slightly affect people's intentions to participate in elections. 2-Interval shows the strongest effects across all measures; 1-Dotplot increases trust the most after elections. Both visualizations create emotional and trust gaps between different partisan identities, especially when a Republican candidate is predicted to win. Our qualitative analysis uncovers the complex political and social contexts of election forecast visualizations, showcasing that visualizations may provoke polarization. This intriguing interplay between visualization types, partisanship, and trust exemplifies the fundamental challenge of disentangling visualization from its context, underscoring a need for deeper investigation into the real-world impacts of visualizations. Our preprint and supplements are available at https://doi.org/osf.io/ajq8f.

The head of the table: marking the "front" of an object is tightly linked with selection.

Objects in the world do not have a surface that can be objectively labeled the "front." We impose this designation on one surface of an object according to several cues, including which surface is associated with the most task-relevant information or the direction of motion of an object. However, when these cues are competing, weak, or absent, we can also flexibly assign one surface as the front. One possibility is that this assignment is guided by the location of the "spotlight" of selection, where the selected region becomes the front. Here we used an electrophysiological correlate to show a direct temporal link between object structure assignments and the spatial locus of selection. We found that when human participants viewed a shape whose front and back surfaces were ambiguous, seeing a given surface as front was associated with selectively attending to that location. In Experiment 1, this pattern occurred during directed rapid (every 1 s) switches in structural percepts. In Experiment 2, this pattern occurred during spontaneous reversals, from 900 ms before to 600 ms after the reported percept. These results suggest that the distribution of selective attention might guide the organization of object structure.

Difficulty limits of visual mental imagery.

While past work has focused on the representational format of mental imagery, and the similarities of its operation and neural substrate to online perception, surprisingly little has tested the boundaries of the level of detail that mental imagery can generate. To answer this question, we take inspiration from the visual short-term memory literature, a related field which has found that memory capacity is affected by the number of items, whether they are unique, and whether and how they move. We test these factors of set size, color heterogeneity, and transformation in mental imagery through both subjective (Exp 1; Exp 2) and objective (Exp 2) measures - difficulty ratings and a change detection task, respectively - to determine the capacity limits of our mental imagery, and find that limits on mental imagery are similar to those for visual short-term memory. In Experiment 1, participants rated the difficulty of imagining 1-4 colored items as subjectively more difficult when there were more items, when the items had unique colors instead of an identical color, and when they scaled or rotated instead of merely linearly translating. Experiment 2 isolated these subjective difficulty ratings of rotation for uniquely colored items, and added a rotation distance manipulation (10° to 110°), again finding higher subjective difficulty for more items, and for when those items rotated farther; the objective measure showed a decrease in performance for more items, but not for rotational degree. Congruities between the subjective and objective results suggest similar costs, but some incongruities suggest that subjective reports can be overly optimistic, likely because they are biased by an illusion of detail.

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.

What Does the Chart Say? Grouping Cues Guide Viewer Comparisons and Conclusions in Bar Charts.

Reading a visualization is like reading a paragraph. Each sentence is a comparison: the mean of these is higher than those; this difference is smaller than that. What determines which comparisons are made first? The viewer's goals and expertise matter, but the way that values are visually grouped together within the chart also impacts those comparisons. Research from psychology suggests that comparisons involve multiple steps. First, the viewer divides the visualization into a set of units. This might include a single bar or a grouped set of bars. Then the viewer selects and compares two of these units, perhaps noting that one pair of bars is longer than another. Viewers might take an additional third step and perform a second-order comparison, perhaps determining that the difference between one pair of bars is greater than the difference between another pair. We create a visual comparison taxonomy that allows us to develop and test a sequence of hypotheses about which comparisons people are more likely to make when reading a visualization. We find that people tend to compare two groups before comparing two individual bars and that second-order comparisons are rare. Visual cues like spatial proximity and color can influence which elements are grouped together and selected for comparison, with spatial proximity being a stronger grouping cue. Interestingly, once the viewer grouped together and compared a set of bars, regardless of whether the group is formed by spatial proximity or color similarity, they no longer consider other possible groupings in their comparisons.

The Unmet Data Visualization Needs of Decision Makers Within Organizations.

When an organization chooses one course of action over alternatives, this task typically falls on a decision maker with relevant knowledge, experience, and understanding of context. Decision makers rely on data analysis, which is either delegated to analysts, or done on their own. Often the decision maker combines data, likely uncertain or incomplete, with non-formalized knowledge within a multi-objective problem space, weighing the recommendations of analysts within broader contexts and goals. As most past research in visual analytics has focused on understanding the needs and challenges of data analysts, less is known about the tasks and challenges of organizational decision makers, and how visualization support tools might help. Here we characterize the decision maker as a domain expert, review relevant literature in management theories, and report the results of an empirical survey and interviews with people who make organizational decisions. We identify challenges and opportunities for novel visualization tools, including trade-off overviews, scenario-based analysis, interrogation tools, flexible data input and collaboration support. Our findings stress the need to expand visualization design beyond data analysis into tools for information management.

Reasoning Affordances with Tables and Bar Charts.

A viewer's existing beliefs can prevent accurate reasoning with data visualizations. In particular, confirmation bias can cause people to overweigh information that confirms their beliefs, and dismiss information that disconfirms them. We tested whether confirmation bias exists when people reason with visualized data and whether certain visualization designs can elicit less biased reasoning strategies. We asked crowdworkers to solve reasoning problems that had the potential to evoke both poor reasoning strategies and confirmation bias. We created two scenarios, one in which we primed people with a belief before asking them to make a decision, and another in which people held pre-existing beliefs. The data was presented as either a table, a bar table, or a bar chart. To correctly solve the problem, participants should use a complex reasoning strategy to compare two ratios, each between two pairs of values. But participants could also be tempted to use simpler, superficial heuristics, shortcuts, or biased strategies to reason about the problem. Presenting the data in a table format helped participants reason with the correct ratio strategy while showing the data as a bar table or a bar chart led participants towards incorrect heuristics. Confirmation bias was not significantly present when beliefs were primed, but it was present when beliefs were pre-existing. Additionally, the table presentation format was more likely to afford the ratio reasoning strategy, and the use of ratio strategy was more likely to lead to the correct answer. These findings suggest that data presentation formats can affect affordances for reasoning.

Rethinking the Ranks of Visual Channels.

Data can be visually represented using visual channels like position, length or luminance. An existing ranking of these visual channels is based on how accurately participants could report the ratio between two depicted values. There is an assumption that this ranking should hold for different tasks and for different numbers of marks. However, there is surprisingly little existing work that tests this assumption, especially given that visually computing ratios is relatively unimportant in real-world visualizations, compared to seeing, remembering, and comparing trends and motifs, across displays that almost universally depict more than two values. To simulate the information extracted from a glance at a visualization, we instead asked participants to immediately reproduce a set of values from memory after they were shown the visualization. These values could be shown in a bar graph (position (bar)), line graph (position (line)), heat map (luminance), bubble chart (area), misaligned bar graph (length), or 'wind map' (angle). With a Bayesian multilevel modeling approach, we show how the rank positions of visual channels shift across different numbers of marks (2, 4 or 8) and for bias, precision, and error measures. The ranking did not hold, even for reproductions of only 2 marks, and the new probabilistic ranking was highly inconsistent for reproductions of different numbers of marks. Other factors besides channel choice had an order of magnitude more influence on performance, such as the number of values in the series (e.g., more marks led to larger errors), or the value of each mark (e.g., small values were systematically overestimated). Every visual channel was worse for displays with 8 marks than 4, consistent with established limits on visual memory. These results point to the need for a body of empirical studies that move beyond two-value ratio judgments as a baseline for reliably ranking the quality of a visual channel, including testing new tasks (detection of trends or motifs), timescales (immediate computation, or later comparison), and the number of values (from a handful, to thousands).