Research
WhispaWave isn’t a hunch. That speaking is faster, keeps your head clear, and that spoken language needs cleaning up is well studied. Here’s the research, linked and verifiable.
We only link real, peer-reviewed or citable sources. No invented quotes.
up to3×faster than typing
extrapolation: ~150 spoken vs. ~40 typed words/min., studies linked below
The evidence
Speaking is measurably faster than typing: not as marketing, but in controlled experiments.
Words per minute
In a controlled smartphone comparison, English via voice was ~3× faster than the keyboard, with about 20% fewer errors.
Ruan, Wobbrock, Liou & Ng (Stanford / UW / Baidu) (2016) Speech Is 3× Faster than Typing for English and Mandarin Text Entry on Mobile Devices · Stanford / arXivPeople comfortably speak at 150–160 words per minute; average typing transcription sits closer to ~40 WPM (handwriting even lower).
Karat et al. (zit. n. Human Factors International) (1999) Human interaction speeds — speaking vs. typing vs. handwriting · CHI / Human Factors InternationalTyping ties up working memory. Speaking leaves more headspace for the actual thought.
The low-level mechanics of writing demand more cognitive resources than speaking, and this remains measurable into adulthood.
Bourdin & Fayol (2002) Even in adults, written production is still more costly than oral production · International Journal of Psychology 37(4)Transcription (typing/writing) consumes working memory and diverts resources from composing and planning. The more automatic it gets, the more headspace is freed for content.
McCutchen; Olive; Kellogg (Review) (2021) The Role of Working Memory in Writing Processes · Frontiers in Psychology (Review)Every switch and interruption has a price: otherwise speed comes at the cost of stress.
Interrupted work wasn’t done slower, but with higher workload, more stress and more frustration. Speed otherwise comes at a steep price.
Mark, Gudith & Klocke (2008) The Cost of Interrupted Work: More Speed and Stress · ACM CHISpoken language is naturally “messy”. Turning it into clean text is the real work.
Spoken
“uh so we meet tomorrow — no, the day after at three”
Clean
Day after tomorrow at 3 pm.
Filled pauses and self-repairs are a regular, built-in part of spontaneous speech. It’s exactly this “mess” that software has to recognize and clean up.
Clark & Fox Tree (2002) Using uh and um in spontaneous speaking · Cognition 84, 73–111Google’s EMNLP benchmark states that discontinuities in spoken transcripts “hamper human readability”. It defines transcript cleanup as its own NLP task. That is exactly what the polish does.
Shen, Zayats, Rocholl, Walker & Padfield (Google) (2023) MultiTurnCleanup: A Benchmark for Multi-Turn Spoken Conversational Transcript Cleanup · EMNLP 2023More sources
The evidence is clear, the daily difference even more so.
No credit card
