Contributor: Alan Davies

ID: 10

Experiment type: Experimental Assay

Technology type: Technology type

Project: GazeMod

Study: 1 hidden item

In this experiment authors we given a set of tasks which they have to accomplish in the lab.

Contributor: Markel Vigo

ID: 3

Experiment type: Experimental Assay

Technology type: Initial rate experiment

Project: whatif

Study: Identifying ontology authoring patterns

Statistical analysis of the relationship between the accessibility of websites and the experience of users with those websites. The accessibility of websites was rated by participants (perceived web accessibility after-use, PWA) and also assessed by different accessibility evaluation methods, that resulted in several accessibility indexes, AI-s. The experience of users was captured by means of post-interaction questionnaires based on the UX model proposed by Hassenzahl.

Contributor: Amaia Aizpurua

ID: 9

Experiment type: Experimental Assay

Technology type: Technology type

Project: Web Accessibility and User Experience

Study: 1 hidden item

Processing big quantities of low-level events and making sense out of them is not a trivial task. We grouped the events in episodes using an empirically solid inactivity threshold, and we used MapReduce to analyse them, using both the user and episode as the "mapping" function. The algorithms create new collection in the database with the resulting behaviours.

Each file serves to analyse different type of behaviours:
MouseBehaviourMongoSkinny.js: different behaviours related with mouse behaviour:

Freely captured low-level data is noisy, and requires a lot of preprocessing before any analysis can be performed.

The most relevant file is ScratchStart.sh. It's a script that is to be executed to process raw data. The scripts need to be executed in that precise order, as some preprocessing scripts have dependencies.

As future work, I might add an intermediate step that deletes all non-analysed users. This step is only necessary if space is needed although I found that discarded users do not

JavaScript injection tool. It's deployed in every Web page in a Web site by adding JavaScript code to them, and sends all interaction data from users to the server.

Copyright (c) 2008 - 2014 The University of Manchester and HITS gGmbH