Design and evaluation of visualizations for large value ranges in multiple time series

This paper investigates the complex issue of visualizing large value ranges in multiple time series. We propose the design spaces for this composed visualization. In two crowdsourced user studies, we test seven designs: Three state-of-the-art designs, three extensions to existing designs, and one novel design. We assess five tasks: Maximum and minimum identification, value discrimination, difference estimation, and slope assessment. Our results show novel findings: For the minimum task, where values in low orders of magnitude have to be identified, our novel height-stack line chart yields the best results. For slope assessment and all tasks where the maximum value can be used as a proxy for the correct answer (maximum, discrimination, and estimation), the linear line graph shows comparable results to all other designs. Moreover, the use of visual mapping to color supports the perception of mantissa and magnitude variations. Unexpectedly, our results indicate that increasing the number of time series does not generally reduce the accuracy of estimation, discrimination, and identification. Our findings are domain independent. They provide useful insights for designers seeking to visualize large value ranges in multiple time series, for example, for financial, medical, or meteorological data.