Knowledge acquisition with domain experts on the aspects of use of visual variables in the Space Time Cube

The Space – Time Cube (STC) is a visual representation developed at the end of the 20th century for understanding the spatio-temporal aspects in human’s everyday life (Hägerstrand, 1970). Since its introduction, it has been widely used in a various discipline (Kraak, 2003; Demšar and Virrantaus, 2010; Tominski, et al., 2012) for the analysis of complex movement data. Despite of increased use, only little is known about its potential from the usability perspective. Previous usability studies (Kristensson, et al., 2009; Demissie, 2010; Kjellin, et al., 2010) focused on the comparison of the STC to other visual representations. However, the usability metrics of the cartographic design theory for the STC content still remain to be unexplored. Therefore, this study particularly focused on the evaluation of the cartographic design aspects into the STC. This study was conducted in two different experiments, but due to the content capacity the abstract will focus on first focus group discussion that was founded on group discussion with experts from geovisualization domain.
The focus group is an informal qualitative method for interactive discussion with experts. It implies group interview, where participants are selected purposefully based on the specific criteria in order to say something on the topic that has to be discussed (Nielsen, 1993). Accordingly, the main objective for focus group interview was to discuss the use of the visual variables based on the cartographic design theory (Bertin, 1983; MacEachren, 1995) with the experts having the knowledge in a geovisualiza-tion domain. This discussion also was the starting point for the establishing the effec-tive design guidelines for the STC content. Of course the preceded stage for it was the investigation of existing STC applications to define the trends on the use of the visu-al variables regarding to the data complexity. But, also the rules of cartographic de-sign theory (Bertin, 1983; MacEachren, 1995; Ware, 2008) in existing textbooks as well as the strategy for data visualization (Shneiderman, 1996). These studies resulted in the list of the hypothesis incorporating existing options of the design with those derived from cartographic textbooks and linked to the overview  detail approach. Consequently, this design options were the objective of the discussion and evalua-tion.
Nine participants attending the workshop on the ‘analysis and visualization of movement data’ at University of Twente had participate for this discussion. The pur-pose of the workshop was to exchange and discusses the experiences on the visualiza-tion and analysis of spatio-temporal data in the STC. As a continuation of the work-shop, participants were invited at the spatial decision room that is equipped with interactive touch screen tables and interactive whiteboard to discuss the usability as-pects of the visual variables and depth cues regarding to the data complexity, Figure 3.
For the evaluation of the visual variables we have selected the use case study with complex movement dataset. Then, applied on it several design options depending on the nature (qualitative / quantitative) of the data. Besides, were hypothesized the effec-tiveness and efficiency of proposed design options for the STC content, also depend-ing on data complexity at overview and zoom steps of the ‘Information Seeking Man-tra’(Shneiderman, 1996). Totally, have been prepared twelve different maps designed in Ilwis open source of STC plugin environment. The difference in qualitative and quantitative nature of the information was considered when dividing materials for evaluation on two touch screens.
The prepared touch screens were running on the windows workstation and allowing direct manipulation of the displayed information content for zoom in/out, rotation and panning options. The discussion session lasted within one hour and aimed to gain insight the aspects of the use of visual variables in the STC. The domain experts as evaluators of visual variables play key role to formulate the hypothesis on the use of the design complexity regarding to the data complexity. Besides, were discussed their influence on the information exploration process. With such approach we could clarify the phases and terms of the use of the visual variables, i. e. under what cir-cumstances are users able to use different design options, and particularly extended design options. This empirical study was only the predecessor step for the next stage, what intend to continue further investigation of design aspects for the STC content. Next evaluations phase looked on it from task based evaluation experiment with non-expert versus users. The combination of different evaluation methods (eye tracking, thinking aloud, screen / video / audio recordings, etc.) will help to reveal more details and answer questions related to the usability metrics.