AS BIG AS A TREE: THE IMPACT OF SCALING CUES AND INTERACTIVE IMAGERY ON VIEWERS' SCALE ESTIMATES, WITH IMPLICATIONS FOR THE USE OF OUTCROP IMAGES AND VISUALIZATION IN GEOSCIENCE EDUCATION

Earth scientists often use images to communicate scientific concepts, commonly providing information to establish the scale of features shown (‘hammer for scale,’ etc.). How effective are these scaling cues? Do observers translate that information correctly, particularly on complex images (e.g., large outcrops with significant foreshortening)? A visualization test was created in which participants were asked to estimate the size of several boxes shown in outcrop photos. All test subjects first viewed a static image, followed by an interactive (gigapan) image of the same outcrop; two different outcrops of two different sizes were used. A control group was asked to estimate the size of features without any scaling information, whereas a second test group was given some general scaling cues, and then asked to perform the same tasks using the same images. Participants (initial test group n=27) represent a range of experience and education levels. Interestingly, the group that was given scaling cues estimated higher (and generally further from correct) than the control group in all cases, but also tended to show more spread in their answers and overall was less likely to be correct than the control group that was not given any scale cues. The control group not only gave smaller size estimates on the static images, but also tended to decrease their estimates during the second ‘interactive’ part of the test. This control group also reported smaller error estimates and showed less spread in the answers (which were generally closer to correct), possibly implying that they were more confident in their answers than the test group. Both groups were closer to ‘correct’ in the second versus first part of the test, which we mainly ascribe to the benefits of being able to interact with the image, although a second-exposure ‘training’ effect cannot be discounted. Results suggest that scale estimates are challenging regardless of experience level, use of scaling cues, etc., and that additional time or interactivity may be needed to successfully communicate features shown on outcrop imagery. Scaling estimates are more difficult for larger/more distant outcrops, and scaling cues may in fact confuse more than they clarify.