Using analytic network process for evaluating mobile text entry methods

Area of Science:

• Human-Computer Interaction
• Usability Engineering
• Decision Science

Background:

• Traditional text entry method evaluations focus on speed and accuracy.
• User preference is a critical but often overlooked factor in selecting text entry methods.
• Existing methods lack flexibility in handling interdependencies between evaluation criteria.

Purpose of the Study:

• To introduce a novel preference evaluation methodology for touch keyboard text entry methods.
• To incorporate user preferences into the selection process of text entry methods.
• To develop a flexible decision model reflecting real-life conditions for text entry method selection.

Main Methods:

• Analytic Network Process (ANP) was employed for preference evaluation.
• A case study involved experts developing a selection decision model for five text entry methods.
• Sensitivity analysis was conducted using simulations to ensure result reliability.

Main Results:

• The QWERTY method was identified as the most preferred text entry method among experts.
• Key arrangement emerged as the most significant criterion in evaluating text entry methods.
• The ANP approach proved robust, with results remaining reliable under perturbation.

Conclusions:

• A multi-criteria decision approach, incorporating user preference, offers a more comprehensive evaluation of text entry methods.
• The study validates the effectiveness of ANP in complex preference evaluations.
• Findings suggest prioritizing key arrangement and user preference in future text entry method design.