Saliency effect in the creation of new mental models of websites




Research project (cognitive psychology)



One woman show

Ines A. Pereira


When a particular object is misplaced, regarding the user’s established mental model, user delays to track the object, or can even disregard it entirely, in a well know phenomenon called change blindness that has been studied by many cognitive psychologists.

The user, when facing this dissonance between his own mental model and the reality of the object’s position, makes an involuntary transition between the cognitive top-down mechanism associated with mental models to a more bottom-up processing, so as to use his other resources and attentional systems to track the object, turning the attention focus to a spatial location determined by the characteristics of that local stimuli (Eysenck & Keane, 2010), more specifically the objects visual saliency (Sutcliffe & Namoun, 2012). This, obviously, slows down the object detection and recognition than when the object is placed according to the user’s mental model. Under this assumption, this study’s hypothesis is that saliency added to a web object misplaced from its most typical location overrides or eases that detection delay.

Although there are already some evidence that saliency can compensate some user’s mental model dissonance, there aren’t enough scientific studies, applied specifically to web objects, that support this hypothesis objectively.

In this study, saliency of specific objects was manipulated jointly with their location. The method involves recording the speed of web objects detection, regarding its location and extra saliency, with the aid of the SaliencyToolbox for the stimuli creation.

The selection of web page type and related web objects was made in a preliminary study, where only the most viewed online stores for 2014 (in Portugal) were selected, and related web objects whose mental model was shown to be more stable within user’s cognitive representations: shop cart, login and search field. Also, this preliminary study was backed with the results of two previous studies developed by Roth et al. (2010, 2013).

For the main study, the goal is to record the speed of web objects detection, regarding its location (typical/atypical) and extra saliency (present/absent), and for that it was designed an empirical experiment, being Superlab the software selected to gather data.

The results from this study can provide to interaction designers a possible technique when facing systems where objects were modified, so as to facilitate the user's mental models creation ou redesign for that web objects.

This study is still in its design stage, the next step is to execute the study with the subjects, obtain results and according conclusions.

Special thanks to my dear teacher and cognitive psychologist PhD. Paulo Noriega, who helped me designing this experiment.

To know more about the method, resources or experiment design, please download the full document (in portuguese, soon will be available an english version).

My personal contribution for this research

(so far)

Theoretical background and development

Experiment design

Resources planning

SaliencyToolbox implementation and use in MATLAB

Stimulus creation with the aid of SaliencyToolbox

Other Research

all research
Go to top