Word Clarity as a Metric in Sampling Keyboard Test Sets
summary
Defines a probabilistic per-word word clarity metric derived from Gaussian touch noise and key geometry, shows clarity explains up to 26.4% difference in error rate and ~25% difference in speed across simulations and a user study, finds smaller keyboards amplify the effect, and produces Pareto-optimized test sets that trade off clarity, bigram representativeness, and memorability; publishes the sampled presets and provides formulas, simulation code and empirical comparisons against random McKenzie sampling.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:31Evaluating keyboard performance
- 3:58Word clarity concept
- 5:01Quantifying word clarity
- 8:14Effect on text entry
- 9:39User study validation
- 11:06Optimizing test phrases
- 14:17Summary and limitations
- 15:21Q&A
related talks
05mCatch Phrase (Testing Edition)
01h 06mKeynote - THE MISMEASURE OF SOFTWARE
46mTalking the Talk: Measuring Chatbot Accuracy
35m"You Can't Miss What You Can't Measure"
33mTL;DR: Testing Copy
44mWhat Playing Wordle Can Teach Us About Web Testing
56mWill your program still be correct next year?
05mCatch Phrase (Testing Edition)