The science of speaking

What the research actually says about getting better at speaking.

One peer-reviewed paper per post, told chart-first. The counter-intuitive findings, the real numbers, and what each one changes about how Speech Away coaches you - drawn from a synthesis of ~50 studies across five literatures.

Figure · Meta-analytic effect sizeA map of the effect sizes

The same intervention type can land anywhere from trivial to transformative. Specific, information-rich, self-referenced feedback clusters high; generic praise clusters low. Each post unpacks one of these results.

Data table

Item	effect size (Cohen's d / Hedges g)
High-information feedback (Wisniewski 2020)	0.99
Treating public-speaking fear (Ebrahimi 2019)	0.74
Self-referenced progress framing (feedback synthesis)	0.55
"A little" feedback (Hattie 2007)	0.39
Praise / low-information feedback (Wisniewski 2020)	0.24

The science of feedback

When a comment helps, when it quietly harms, and what separates the two.

≈1 in 3 About one in three feedback interventions makes performance worse Kluger, A. N., & DeNisi, A. (1996) Read the breakdown → 4 levels Where you aim feedback matters more than how much you give Hattie, J., & Timperley, H. (2007) Read the breakdown → d = 0.99 High-information feedback works four times better than praise Wisniewski, B., Zierer, K., & Hattie, J. (2020) Read the breakdown →

How practice becomes skill

Why the design of practice - not the amount - decides what transfers to the real moment.

50% The feedback that makes you better today can stop you learning Salmoni, A. W., Schmidt, R. A., & Walter, C. B. (1984) Read the breakdown → crossover The same guidance that helps a beginner holds back an expert Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003) Read the breakdown → proximal > distal Small, near goals build confidence; one distant goal does not Bandura, A., & Schunk, D. H. (1981) Read the breakdown →

Motivation, mindset & nerves

How praise, expectations, and anxiety shape whether you keep going.

trait praise ↓ Praise the move, not the talent - or the next stumble undoes them Mueller, C. M., & Dweck, C. S. (1998) Read the breakdown → 27% → 64% One sentence of belief more than doubled who acted on criticism Yeager, D. S., et al. (2014) Read the breakdown → g = 0.74 Speaking anxiety is common - and it responds strongly to practice Ebrahimi, O. V., Pallesen, S., Kenter, R. M. F., & Nordgreen, T. (2019) Read the breakdown →

What listeners actually hear

The delivery features that move perception - and the ones everyone overrates.

140-160 wpm Faster speech persuades - until it doesn't Miller, N., Maruyama, G., Beaber, R. J., & Valone, K. (1976) Read the breakdown → ≤ 5 / min Fillers don't hurt until you cross a threshold - and "uh" can help Laske, M., et al. (2024) Read the breakdown → ≈1% Vocal fry is judged negatively - by about one percent Anderson, R. C., Klofstad, C. A., et al. (2014) Read the breakdown → ~30% The most powerful delivery lever is the one a transcript can't see Guyer, J. J., Fabrigar, L. R., & Vaughan-Johnston, T. I. (2019) Read the breakdown →

Can a machine coach you?

What automated speech-coaching trials found about feedback that actually teaches.

N = 90 A virtual coach worked - but only the personalized post-hoc report Hoque, M. E., Courgeon, M., Martin, J.-C., Mutlu, B., & Picard, R. W. (2013) Read the breakdown → p = 0.010 Live feedback keeps you engaged; the delayed report is what teaches Chollet, M., Marsella, S., & Scherer, S. (2022) Read the breakdown →