Information visualizations for knowledge acquisition: The impact of dimensionality and color coding

https://doi.org/10.1016/j.chb.2005.01.006Get rights and content

Abstract

Information visualizations – interactive graphical representations of large amounts of abstract data which do not have a natural visual representation – have mainly been used to support information retrieval. This article investigates whether information visualizations are also suitable for fostering knowledge acquisition as well as how information visualizations, from a cognitive perspective, have to be designed to be efficient learning tools. An experimental study provided evidence that information visualizations support knowledge acquisition. In addition, with regard to the appropriate design, the empirical results showed that two-dimensional information visualizations are better suited for supporting processes of knowledge acquisition than three-dimensional ones and that color-coded information visualizations slightly increase performance in a knowledge test compared to monochromatic ones.

Introduction

Technological innovation enables the storing of fast-growing quantities of information. Accordingly, it has become increasingly important to develop efficient methods to structure large and complex information sets. Recently, there have been several attempts to tackle this challenge by using computer-based information visualizations, that is, graphical representations of large amounts of abstract data which do not have a natural visual representation (Wiss, Carr, & Jonsson, 1998). According to Card, Mackinlay, and Shneiderman (1999), information visualizations can be characterized as “computer-supported, interactive, visual representations of abstract nonphysically based data to amplify cognition” (p. 6). For instance, information visualizations have been used to display the information units of abstract data sets like document collections or text-based information contents of the World Wide Web. A simplified sketch of the type of spatial information visualization used in the empirical study presented in this article is shown in Fig. 1. In this sketch, four information units (A, B, C, and D) are displayed together with three of their attributes that are represented by means of three spatial dimensions. In other words, the four information units are arranged in a three-dimensional information space according to their values for the three attributes that are represented by three orthogonal axes.

Information units pool those parts of data sets that belong together. The units can be described by their values of numerous different attributes. Typically, only a subset of these attributes can be represented spatially (i.e., three attributes at the most). Thus, other attributes of the information units may be represented textually (e.g., by means of a pop-up window) or by other representational codes (e.g., color coding).

Currently, there is a trend in computer science to develop technically complex information visualizations. This trend – which is motivated by the intention to include the current state-of-the-art technologies – has led to an increased development of information visualizations that use three spatial dimensions to visualize data. Research with regard to these advanced information visualizations has mainly been focusing on solving technical problems in the context of information-retrieval tasks where information visualizations prove very helpful in enhancing a user’s ability to find and apply information. However, from the perspective taken here there are at least two shortcomings with regard to the current research on information visualizations in computer science:

  • First, information visualizations may not only facilitate information retrieval and application, but may also enhance the recognition of structures and relationships within the data visualized. Therefore, the question arises as to whether information visualizations are also useful to support knowledge acquisition from the presented data and their inherent structure. However, so far, this question has not yet been investigated.

  • Second, a consideration of the user’s perspective in terms of cognitive processes and limited cognitive resources is lacking in most research on information visualizations. In particular, little is known about the cognitive processes involved in – and perhaps supported by – the use of information visualizations as learning tools.

Therefore, the aim of the empirical study presented in this article was to investigate to which extent multi-dimensional information visualizations are superior to a non-spatial representation when the task is to memorize a data set and to acquire an understanding of the relationships embedded within this set. Moreover, in the study it was investigated how to design information visualizations in order to promote learning outcomes. First, it was investigated experimentally whether information visualizations should be two-dimensional or whether a third spatial dimension for information representation may be helpful for knowledge acquisition. Second, the question was addressed as to whether knowledge acquisition with spatial information visualizations can be further enhanced by using color coding to represent attributes of data.

Section snippets

What is the pedagogical potential of information visualizations?

There are different cognitive theories that can be used to argue that information visualizations may be efficient tools to enhance the acquisition of knowledge from large and abstract data sets, whereby knowledge acquisition refers to understanding and memorizing abstract data and their interrelations.

First, theories of computational efficiency pay specific attention to the inferences learners have to make in order to understand a task or a domain. The argument here is that some

How to design information visualizations for knowledge acquisition?

Beyond the general claim that information visualizations are tools that might foster the acquisition of knowledge from large and abstract data sets, the research question is how to design information visualizations that foster learning outcomes in particular. Thus, the study presented in this article addresses how the dimensionality of information visualizations and color coding of attributes might affect learning.

Experimental environment

The experimental environment used in this study was developed in the context of the European project “Mummy” of the Computer Graphics Center in Darmstadt (Germany), which focuses on mobile knowledge management using multimedia-rich portals for context-aware information processing, for instance, at construction sites. In particular, the experimental environment was designed to provide architects with an overview of the details of their construction projects. Each project is described by values

Experiment

This experiment first investigated whether information visualizations are more suited to foster knowledge acquisition than text-based information representations. It then analyzed whether dimensionality and color coding of information visualizations influence learning.

Summary and conclusion

This experiment provided evidence for the suitability of information visualizations for knowledge acquisition. These findings are in line with theories of computational efficiency (Larkin & Simon, 1987), the cognitive theory of multimedia learning (Mayer, 2001), assumptions concerning the existence of a visuospatial working memory (Logie, 1995) and cognitive load theory (Sweller et al., 1998). According to these theories, information visualizations may be more computationally efficient as

Acknowledgement

We thank Matthias Grimm of the Computer Graphics Center Darmstadt (Germany) for programming the information visualizations.

References (25)

  • M. Hicks et al.

    Comparison of 2D and 3D representations for visualising telecommunication usage

    Behaviour & Information Technology

    (2003)
  • T. Huk et al.

    Computer-animations in education: the impact of graphical quality (3D/2D) and signals

  • Cited by (70)

    • Impacts of cues on learning: Using eye-tracking technologies to examine the functions and designs of added cues in short instructional videos

      2020, Computers in Human Behavior
      Citation Excerpt :

      This pattern shows that the textual cues only promoted the cognitive process of selecting information, whereas the visual cues and the combined textual-&-visual cues promoted the cognitive processes of selecting, organizing and integrating information. This result is consistent with the previous studies, which found that cues tended to promote the multimedia learning effect (Boucheix & Lowe, 2010; De Koning et al., 2010, 2007; Jeung, Chandler, & Sweller, 1997; Keller, Gerjets, Scheiter, & Garsoffky, 2006; Mautone & Mayer, 2001; Ozcelik, Karakus, Kursun, & Cagiltay, 2009). To answer the second question, we compared the impacts of the three different cues on learning outcomes.

    • Occupant perceptions of building information model-based energy visualizations in eco-feedback systems

      2018, Applied Energy
      Citation Excerpt :

      Kalyuga et al. [40] demonstrated that color-coded formats reduced working memory load by reducing cognitive search processes. The results of Keller et al. [41] extended these findings, showing that color-coded information visualization increases comprehension. Similarly, cognitive science fields have found spatial displays also facilitate information searching and organization, freeing up greater space for cognitive processing [42].

    View all citing articles on Scopus
    View full text