Type of research:
Type of stimuli:
The aim of the experiment was to investigate a) the influence of color distance between intervals/classes of sequential and qualitative schemes b) spatial distance between selected areas on their distinguishability.
We systematically evaluated 5 sequential and 5 qualitative colour schemes in a two-stage user experiment. At first, we conducted an online study to obtain data of large variety of heterogeneous population. Following this, in a controlled lab study with eye-tracking, we re-examined the findings from the online study for a subset of experimental stimuli and further assessed the user experience through an analysis of their visual behaviour.
There was a single task: locate two or three areas on the map (marked with a dot), compare them and decide whether these areas were of the same colour.
On each stimulus, two areas were selected and marked with a dot. The colour distances between the selected areas were controlled to be same number of ∆E00= 2, ∆E00= 4, ∆E00= 6, ∆E00=8, ∆E00= 10 or ∆E00=0 (same colour). The spatial distribution of these areas were also controlled; to simulate various conditions that could occur in map reading, we distributed the two areas over the map: they could be next to each other, at mid-distance on the map, or at two extremes.
In total 106 stimuli were used in the WS and 41 in the ET experiment. The stimuli used in the ET are a subset of the ones from the WS.
211 volunteers have participated in the online experiment (WS) (121 females, 90 males). The controlled lab experiment (ET) had 32 voluntary participants (19 females, 13 males).
For the eye-tracking study the remote eye-tracking device SMI RED 250, developed by SensoMotoric Instruments, was used. This device was operating at frequency of 120 Hz. The web survey was implemented in the open source application LimeSurvey.
To evaluate the effect of the color distance, along with accuracy and response time, eye tracking metrics fixation frequency, fixation duration and scanpath speed as well as a gaze transition analysis were utilized.
The hypotheses of the study were as follows:
- (H1) Increasing colour distance will lead to a more accuracy
- (H2) Increasing colour distance will result in shorter response time
- (H3) Sequential and qualitative colour schemes will have no difference in accuracy or response time at the same colour distance level
- (H4) Smaller colour distances will cause: longer average fixation durations (indicates difficulty in extracting information), higher frequency of fixations (indicates less efficient searching), longer scanpaths (indicates less efficient searching), and an increasing number of revisits between compared areas.
Regardless of the minor differences between the WS and the ET, our results clearly demonstrated that the lowest examined colour distance ∆E00 = 2 led to more errors and slowed people down. We can also safely say that ∆E00 = 10 is a safe colour distance, but the rest of the tested distances are essentially not recommendable. In this particular study, basic eye-tracking metrics (fixations frequency, fixation duration, scanpath speed) did not offer additional explanations.
Overall response time and gaze transition analysis showed that distinguishing between two areas is more difficult on sequential maps, than qualitative.
Brychtova, A. & Coltekin, A. (to be appeared). Discriminating Sequential and Qualitative Colour Schemes
Brychtova, A. & Coltekin, A. (to be appeared). The effect of the distance between map symbols on the discriminability of their colors