Visual display of qualitative information: Implications of using illustrations to depict video data

[This paper is part of the Focused Collection on Qualitative Methods in PER: A Critical Examination.] Qualitative studies in the domain of physics education research have become more common in the last several decades. Methodologically, this has been marked by an expansion of the types of data collected in physics education research (PER): namely, in the use of individual and group interviews, problem-solving sessions, and classroom observations—often in the form of video data. However, depictions of those video data in publications, a key opportunity for communicating analytic approaches and results, are often left out or are constructed without addressing some important implications of video data depictions. In this paper, we critically examine how video data can be depicted, first providing a rough taxonomy of depiction approaches. We then emphasize the case of line illustrations as a depiction approach with significant potential for improving (i) the transparency of analytic perspectives, (ii) communicational clarity, and (iii) the handling of ethical issues, such as participant anonymization, followed by a discussion of the limitations and challenges of using illustrations. In the Appendix, we provide a practical guide for researchers interested in making their own illustrations for academic publishing purposes.

This article appears in the following collection:

Figure 1

Examples of nonsequential depictions of video data, categorized as wide frames, narrow frames, processed frames, and illustrations—all with unannotated and annotated variants. (a) is an AI-generated image resembling the figures in Azevedo and Mann [31] (for copyright purposes). The remaining images are reproduced under a Creative Commons license from the following sources: (b) Hu and Rebello [32], (c) Sabo and Elby [33], (d) redrawn in the style of Alibali and Nathan [34] (for copyright purposes), (e) Frank and Scherr [35], (f) Scherr [5], (g) Gupta et al. [36], and (h) Euler et al. [37].

Figure 2

Examples of sequential depictions of video data, otherwise following the same categorization as Fig. 3. Images are reproduced under the Creative Commons license from the following sources: (a) Phillips et al. [38], (b) Stephens and Clement [39], (c) Robertson and Hairston [28], (d) Stephens and Clement [39], (e) Odden [4], (f) Eriksson et al. [40], (g) Gregorcic et al. [3], and (h) Euler et al. [2].

Figure 3

Examples of the same source frame of video transformed into the various types of video data depictions, following the same categorization scheme as Figs. 1 and 2. The original frame of the video (upper left) was sourced from the supplementary videos of Scherr [5], while the remaining depictions were generated by us for the purposes of this paper.

Figure 4

An example of supplementing an illustration of video data with a closeup of a digital menu and a student’s environmentally coupled gesture. Reproduced from Euler et al. [37] under the Creative Commons license.

Figure 5

Examples of single illustrations (nonsequential) with multiple-exposure-type “action sequences” for depicting the motion of (a) environmental elements (in the simulation, an arrow projectile flies from left to right and gets jammed in a soft object) and (b) gestures. (a) is reproduced from Euler et al. [37] under the Creative Commons license and (b) was redrawn in the style of Alibali and Nathan [34] (for copyright purposes).

Figure 6

Hybrid video data depictions reproduced under the Creative Commons license from Asplund and Kilbrink [57], blending elements of video frames, and illustrations.

Figure 7

Hybrid video data depictions in the style of Vossoughi et al. [62], blending elements of video frames, and illustrations. An AI-generated image was used for copyright purposes.

Figure 8

A stepwise demonstration of how the so-called pen tool works to make Bézier curves. The Bézier curve “handles” are shown with straight lines ending with circles. These handles show the angle and “strength” of the tangent line for the thicker line at that point. (a) A line is previewed from a first point (the left square) toward a second point, with the mouse hovering (not clicked); (b) the mouse is clicked and dragged up and to the right, which causes the drawn line to curve upward into the second point; and (c) the mouse is released, committing the s-curved line between the first two points and starting the process all over.