Sketch maps

\citet{r.1977} defined a cognitive map as ‘‘a set of stored propositions about the environment, each having been assigned a true value by a given individual.’’ Keeping that definition in mind, sketch maps are sorts of reflections of individual cognitive maps. According to \citet{v.2004}, sketch maps are “compact spatial representations that express the key spatial features of a situation for the task at hand, abstracting away the mass of details that would otherwise obscure the relevant aspects”. Either two definitions are similar in terms of the way that they mention the cognitive process with one emphasizing the individual characteristics and the other emphasizing the importance of the task. Therefore, interpretation of sketch maps offers a great chance to gain insight about cognitive maps.
Sketch maps have been used in several research projects as a data collection method for investigating cognitive processes of map users (\citealp{Billinghurst}; \citealp{v.2004} ; \citealp*{Tu_Huynh_2007}; \citealp*{2011}; \citealp*{k.2012}; \citealp{Ooms_2013}). Sketch maps are often combined with think aloud procedure as a complementary data collection method (\citealp{Ooms_2013} ; \citealp{Kettunen_2014}) because thinking aloud gives insights in the user’s unfiltered thoughts. Thinking aloud itself, however, as the disadvantage that also consumes part of the user’s memory capacity.

Retrieving a sketch map from memory

With sketch map evaluation, we intend to examine map users’ cognitive processes of learning, acquiring and remembering the information presented via screen maps via drawing. In order to analyze participants’ cognitive behaviors through sketch maps, we need to identify the cognitive procedures involved during memory task. The first step is orientation of the participant (i.e. establishing a strategy to execute the task from the beginning to the end) and the surroundings (in this case drawing environment and its tools). The second step is task execution, in which participants form links between cognitive processes through WM and LTM. In chronological order, the participant first consults his WM to check whether there is information about map elements that has to be drawn. If the answer is yes, the participant will draw these elements; if no, he has to consult LTM, which is responsible for recalling act. For a participant to draw an element whose information stored in LTM, that information needs to be transferred to WM. Afterwards, evaluation takes part for editing or redrawing and then the participant asks WM once again to finalize the procedure (\citealp{v2015}). It is important to keep in mind that, this procedure is repetitive and continues until the participant is satisfied with the result. During this procedure, sensory memory captures the image of the sketch map and transfer it to the WM. The memories of original stimulus, which was previously stored in LTM, need to be recalled. Once the participant retrieves that information, he is able to compare the sketch map with the original stimulus depending on the location, size, shape, colour, and so on. Retrieval process of chunks of information (LTM-WM) requires activation of the related information. This involves pointers, schemas and links between schemas stored in the LTM. Pointers activates and retrieves the desired chunks of information from LTM and place them in WM (\citealp{Ooms_2013}).
In this paper, we tried to evaluate the above-mentioned cognitive process on a static screen map to find out the cognitive abilities and/or limitations of map users when they on one hand, study the map and on the other hand, retrieve this information later on.

Eye tracking

The recent developments in technology led visual stimuli evolve from analog and static to digital and interactive (\citealp{Ooms_2013a}). Inherently, cartographic products, which are complex visual stimuli, recently are available more in digital forms as screen maps. Although an excessive amount of new cartographic products exists, there is still a little known about their use, user and usability issues. Therefore, we, cartographers need to pay more attention to understanding the attentive behaviors of users dealing with screen maps.
Eye tracking, as a widely-used quantitative user-testing method, has contributed to human-computer interaction usability studies in numerous disciplines varying from psychology to software engineering, marketing, sports, aviation, navigation and so forth (\citealp{Crundall_2002}\citealp{Ball_2003}\citealp{Jacob_2003};  \citealp*{Poole}\citealp{Schriver_2008}\citealp*{Wedel_2006};  \citealp*{f.2009}). Many cartographers also employed eye tracking in their usability research, especially for the assessment of visual elements. Eye tracking can either be applied as a standalone technique, or along with other methods such as think-aloud, questionnaire, and interview (\citealp{_ltekin_2010}\citealp{van_Elzakker_2008}\citealp{Irina_Fabrikant_2008}\citealp{Fabrikant_2010}\citealp{MacEachren};  \citealp*{k.2012}; \citealp{Ooms_2014}; \citealp*{philippe2015}; \citealp{Ooms_2016}).  
Eye tracking data can provide useful statistics related to users’ gaze activity that can be linked to their cognitive processes when they interact with visual stimuli on the screen. It consists of a list of pixel coordinates of the screen regarding to various positions of the gaze (POR: point of regard). Hence, we can derive valuable information from eye tracking statistics such as; how long (fixation duration) and how often (fixation count) a person focuses on a specific area of interest together his scan-path characteristics (the length and speed of the gaze activity) (\citealp*{2007};  \citealp{Ooms_2014}).
In this paper, for evaluation of the performances of participants, we made use of eye tracking statistics such as number of fixation per second, the average duration of fixation, and trial durations for study and drawing time, individually. This will be further discussed in methodology section.

Study Design

Participants

A total of 56 participants took part in the study, with 24 experts and 32 novices. The number of female and male participants was 8 and 24 for novices, and 13 and 11 for experts, respectively. Thus, we had four different user groups categorized as expert females, expert males, novice females and novice males. Novice participants were undergraduate Business and Economy students whose ages varying between 18 and 24 years and who gained credits in return of their participation. The expert group consists of participants who had at least a MSc. in Geography, Geomatics Engineering or related areas and all of them were affiliated to Department of Geography (Ghent University). While experts work with cartographic products on a daily basis; novices neither used cartographic products in their daily life, nor were trained before the experiment. The language of the entire experiment was English.

Stimulus

For this experiment, we selected the stimulus from the Belgian 1:10k topographic map series. We paid attention that it was not too complex, yet contained some specific main structuring elements. To combat the learning affect, the selected map did not belong to a well-known area/city (Figure 2).