Scientific figures, i.e. visuals such as graphs and diagrams, are an important component of Intergovernmental Panel on Climate Change (IPCC) reports that support communication and policy-making. It is therefore imperative that figures are robust representations of the science and are accessible to target audiences. We interviewed IPCC authors (n = 18) to understand the development of figures in the IPCC Fifth Assessment Report (AR5) Working Group 1 (WG1) Summary for Policy-Makers (SPM). Authors expressed the view that the need to maintain scientific accuracy constrained making figures more accessible, with the consequence that figures retained complexity and often required specialists to explain the figures to others. Using sort tasks with IPCC authors and with a group of non-specialists (undergraduate students; n = 38), we found that IPCC authors generally had good awareness of which figures non-specialists perceived as being most difficult to understand. Further, by evaluating the visual complexity of the AR5 WG1 SPM figures using a computational measure, we found that greater visual complexity (i.e. high quantity of information, use of multiple colours and densely packed visual elements) is associated with greater perceived comprehension difficulty. Developing and integrating computational approaches to assess figures alongside user testing could help inform how to overcome visual complexity while maintaining scientific rigour and so enhance communication of IPCC figures and scientific visuals.
22 September 2016 in Cognitive Research: Principles and Implications
Novices struggle to interpret maps that show information about continuous dimensions (typically latitude and longitude) layered with information that is inherently continuous but segmented categorically. An example is a topographic map, used in earth science disciplines as well as by hikers, emergency rescue operations, and other endeavors requiring knowledge of terrain. Successful comprehension requires understanding that continuous elevation information is categorically encoded using contour lines, as well as skill in visualizing the three-dimensional shape of the terrain from the contour lines. In Experiment 1, we investigated whether novices would benefit from pointing and tracing gestures that focus attention on contour lines and/or from three-dimensional shape gestures used in conjunction with three-dimensional models. Pointing and tracing facilitated understanding relative to text-only instruction as well as no instruction comparison groups, but shape gestures only helped understanding relative to the no instruction comparison group. Directing attention to the contour lines may help both in code breaking (seeing how the lines encode elevation) and in shape inference (seeing how the overall configuration of lines encodes shape). In Experiment 2, we varied the language paired with pointing and tracing gestures; key phrases focused either on elevation information or on visualizing shape. Participants did better on items regarding elevation when language highlighted elevation and better on items requiring shape when language highlighted shape. Thus, focusing attention using pointing and tracing gestures on contour lines may establish the foundation on which language can build to support learning.
1 January 1998 in Journal of Experimental Psychology: Human Perception and Performance
New phenomena and results are reported that implicate a common contour interpolation mechanism in illusory and occluded (modal and amodal) object completion. In 3 experiments, a speeded classification task was used to study novel quasimodal displays in which occluded and illusory contours join. Results showed the same advantages in speed and accuracy over control displays for quasimodal, illusory, and occluded displays. The implications of quasimodal displays, along with another new display type in which contour linkages must precede determination of modal or amodal appearance, are considered. These logical considerations and empirical results suggest that amodal and modal completion depend on a common underlying mechanism that connects edges across gaps.
13 June 2019 in Cognitive Research: Principles and Implications
Research examining the relation between spatial skills and the science, technology, engineering and mathematics (STEM) fields has focused on small-scale spatial skills, even though some STEM disciplines—particularly the geography and geoscience (GEO) fields—involve large-scale spatial thinking at the core of their professional training. In Study 1, we compared large-scale navigation skills of experienced geologists with those of experienced psychologists, using a novel virtual navigation paradigm as an objective measure of navigation skills. In Study 2, we conducted a longitudinal study with novice Geographic Information Systems (GIS) students to investigate baseline navigational competence and improvement over the course of an academic semester. In Study 1, we found that geologists demonstrated higher navigational competence and were more likely to be categorized as integrating separate routes, compared to their non-STEM counterparts. In Study 2, novice GIS students showed superior baseline navigational competence compared to non-STEM students, as well as better spatial working memory and small-scale mental rotation skills, indicating self-selection. In addition, GIS students’ spatial skills improved more over the course of a semester than those of non-STEM students. Our findings highlight the importance of large-scale spatial thinking for enrollment and success in the GEO fields but likely also across the broader range of thinking involving spatial distributions. We discuss the potential of GIS tools to develop spatial skills at an early age.