Scanning Keyboards Are Computer Based Communication Aids Used By Individuals With Speech And Motor Disorders. A Scanning Keyboard Refers To A Soft Keyboard Operated With A Scanning Input Method. The Traditional Design Method For These Keyboards, Which Involves Testing Of Prototypes With Physically Disabled Users, Is Problematic As It Is Difficult To Get Users Or Collect Data. In This Thesis, An Alternative Design Approach Is Proposed To Overcome This Problem. In The Proposed Design Method, Models Are Used To Compute Performance Of Scanning Keyboard Users In Terms Of Text Entry Rate. Designs Are Compared On The Basis Of The Computed Text Entry Rates. Validity Of The Proposed Models Has Been Demonstrated With Results From Empirical Study. The Text Entry Rate Models Are Developed Assuming Error Free User-Computer Interaction. In Practice, However, It Has Been Found That Errors During Input Selection Affect Performance Of Scanning Keyboard Users Significantly. To Account For User Errors In Scanning Keyboard Design, An Error Measure Called The “Error-Proneness” Or Ep Has Been Proposed. A Model Has Been Developed To Compute Ep. In Order To Develop The Computational Error Model, Data On Scanning Keyboard User Errors Have Been Collected Through An Empirical Study. From Empirical Observations, A State Transition Model (Stm) Of User Behavior Has Been Developed, Which Could Explain Errors. The Stm Has Been Used To Predict Distributions Of Error Probabilities With Respect To The Distance Between Keys. These Distributions Form The Basis Of The Model For Ep Computation. Furthermore, A Set Of Design Principles Has Been Derived From These Distributions. On The Basis Of The Two Measures, Namely The Text Entry Rate And Ep, Two Design Approaches Have Been Proposed. One Of These Approaches Can Be Used To Determine The Best Design From A Set Of Design Alternatives. The Approach Is Suitable When There Is Limited Number Of Design Alternatives. In The Other Design Approach, A Dual Objective Simulated Annealing Algorithm Is Used To Explore The Scanning Keyboard Design Space, In Order To Identify A Good Design. An Empirical Study Has Been Carried Out To Show The Efficacy Of The Proposed Design Approaches. Keywords: Augmentative Communication, Soft Keyboards, Scanning And Access Switches, Fitts’-Digraph Model, Text Entry Rate, Focus Movement And Selection Time, Focus Distance, Timing Errors, Selection Errors, Error-Proneness, Design Space Exploration, Key Grouping, Dual Objective Simulated Annealing.