Not a Flaw: Is Speaking Anxiety Inborn? Beatty, McCroskey & Heisel (1998)
By Noam Shemla · The Studies
TL;DR For decades, fear of speaking was treated as something people learn. Michael Beatty, James McCroskey and Alan Heisel argued the opposite: that it is mainly an expression of inborn temperament, a blend of introversion and neuroticism, two traits they took to be 80% inherited. Their own data showed the blend is real. The 80% did not hold up, and the conclusion they drew from it, that treatment can do little, has been contradicted since.
- Paper: Beatty, M. J., McCroskey, J. C., & Heisel, A. D. (1998). Communication apprehension as temperamental expression: A communibiological paradigm. Communication Monographs, 65(3), 197–219.
- Kind: A theoretical essay with one small supporting study, not an experiment or a genetic study.
- Data: 73 randomly selected adults (42 women, 31 men) completed McCroskey’s 24-item speaking-anxiety scale and Eysenck’s short scales for extraversion and neuroticism.
- Claim: Communication apprehension is mainly the expression of inherited temperament, and learning plays a small part in it.
- Access: Read in full from the scanned copy on James McCroskey’s own publications site.
Communication apprehension, the fear of talking to others that includes the fear of speaking to an audience, was defined by James McCroskey in the 1970s, and for twenty years it was explained mostly as learned: a child punished or ignored for speaking learns to fear it. In this essay, McCroskey, who had written that view himself, argued with two colleagues that the explanation was mostly wrong.1 They called the approach communibiology: the view that differences in how people communicate come largely from differences in their brains, and those largely from their genes.
The argument in three steps
The case rests on a chain. Speaking anxiety is a blend of two broad personality traits. Those traits are mostly inherited. So speaking anxiety is mostly inherited too.
The link they measured: anxiety and temperament
To show that speaking anxiety is really introversion plus neuroticism under another name, the authors gave 73 randomly selected adults all three questionnaires, in random order. The overlap was very large.
How speaking anxiety related to the two traits (73 adults)
| Value [95% CI] | Corrected | |
|---|---|---|
| Speaking anxiety × introversion | .83 [.74, .89] | .90 |
| Speaking anxiety × neuroticism | .74 [.61, .83] | .80 |
| Introversion × neuroticism | .67 [.52, .78] | |
| Variance explained | R² = .75 | R = .93 |
| Weight of introversion | β = .60 | |
| Weight of neuroticism | β = .34 |
Every number in that table can be rebuilt from the three correlations, which is a useful check on a scanned 1998 paper. It passes.
The overlap is striking, but it comes with a caveat the paper does not dwell on. All three were self-reports, filled in by the same people at one sitting, and the speaking-anxiety scale and the neuroticism scale both ask about feeling nervous. Some of the overlap is the same feeling measured twice.
The link they quoted: temperament and genes
The heritability figure came from elsewhere. Citing Hans Eysenck's reviews of twin research, the authors wrote that the ratio of genetic to environmental influence on introversion and neuroticism is about 80/20, and that better measurement would only push it further toward genes.
Heritability has a precise meaning, and it is worth being clear about it. In a twin study, it is estimated by comparing how alike identical twins are, who share all their genes, with how alike fraternal twins are, who share about half.
Two things follow. Heritability describes differences across a population, not a person: 80% does not mean that 80% of your nerves are fixed. And it says nothing about whether a trait can change. A trait can be highly heritable and still respond to what happens to you.
The 80% itself was generous. Pooling the behaviour-genetic studies of personality since, a 2015 meta-analysis put the heritability of personality at about 40%.2 Twin studies, the kind Eysenck relied on, give higher figures than family and adoption studies, a pattern the meta-analysis confirmed.
The 65% needs a word of context. It comes from Beatty and colleagues' own meta-analysis of twin studies, four years later, which pooled research on social anxiety and found it 65% heritable, alongside 70% for interpersonal affiliation and 58% for aggressiveness.3 That supports a substantial genetic part in social anxiety. It rests on twin designs, which the 2015 analysis found give the higher estimates.
What would the chain imply for speaking anxiety itself? A rough illustration, of our own: if the traits explain 75% of speaking anxiety, and the traits' genetic share carried straight through, the inherited part of speaking anxiety would be about 0.75 × h².
Even on the paper's own inputs, the chain delivers something nearer 60% than 80%. On today's figures for personality it delivers about a third. Speaking anxiety is partly inherited; "mostly" is the part of the claim that has not held.
The claim about treatment
The essay went further. If speaking anxiety is mostly temperament, the authors argued, treatment should have little room to work. They did the arithmetic on their own scale: the 24-item measure runs from 24 to 120, with a standard deviation of 15.3, so even a large effect of half a standard deviation would move someone by fewer than eight points and leave a very anxious speaker still very anxious.
This is the part of the paper later evidence tested most directly. A 2019 meta-analysis pooled 30 randomised controlled trials of psychological treatments for fear of public speaking, with 1,355 participants.4
Fear of public speaking, in other words, responds to treatment about as well as most things psychology treats, and the effect lasts. The authors had allowed for this, calling it an anomaly that environment would have to explain. The anomaly turned out to be large.
What this paper cannot tell you
- How heritable speaking anxiety is. The paper measured no genes and studied no twins. Its 80% was quoted, for other traits.
- Why the traits and the anxiety overlap. Correlations in one sample of 73 people, from self-reports taken together, show overlap, not the direction of cause.
- Anything about a particular person. Heritability describes a population, and even a high figure leaves room for change.
- Whether the debate is over. It is not. Celeste Condit argued that the model left too little room for culture and learning, and the authors replied in the same journal issue.5
What a speaker can take from it
The part that holds up is the kind one. If speaking frightens you more than it seems to frighten other people, that is not a failing of character or effort. Temperament plays a real part, and some people start with more fear than others.
The part that has not held up is the idea that where you start is where you stay. The fear falls with repetition, as the exposure research shows, and it shows less than you think, as the illusion of transparency shows. Treatment trials show the gains can be large and lasting. A temperament is a starting point, not a verdict.
This article is built from the full text of Beatty, McCroskey & Heisel (1998), read from the scan on McCroskey’s site, and from the published abstracts of Beatty et al. (2002), Vukasović & Bratko (2015) and Ebrahimi et al. (2019). The correlations, weights and reliabilities are the paper’s own. The intervals, the reconstruction of the regression, the fully corrected R, the chain illustration and the conversion of treatment effects into scale points are our calculations.
Measured
- Speaking anxiety, introversion and neuroticism, by self-report, in 73 adults
Inferred, not measured
- That speaking anxiety is mostly inherited: an argument built on a quoted 80/20 figure, not measured in the paper
- That treatment can do little: a prediction, since contradicted by randomised trials
References
- Beatty, M. J., McCroskey, J. C., & Heisel, A. D. (1998). Communication apprehension as temperamental expression: a communibiological paradigm. Communication Monographs, 65(3), 197–219. https://doi.org/10.1080/03637759809376448
- Vukasović, T., & Bratko, D. (2015). Heritability of personality: a meta-analysis of behavior genetic studies. Psychological Bulletin, 141(4), 769–785. https://doi.org/10.1037/bul0000017
- Beatty, M. J., Heisel, A. D., Hall, A. E., Levine, T. R., & La France, B. H. (2002). What can we learn from the study of twins about genetic and environmental influences on interpersonal affiliation, aggressiveness, and social anxiety? A meta-analytic study. Communication Monographs, 69(1), 1–18. https://doi.org/10.1080/03637750216534
- Ebrahimi, O. V., Pallesen, S., Kenter, R. M. F., & Nordgreen, T. (2019). Psychological interventions for the fear of public speaking: a meta-analysis. Frontiers in Psychology, 10, 488. https://doi.org/10.3389/fpsyg.2019.00488
- Condit, C. M. (2000). Culture and biology in human communication: toward a multi-causal model. Communication Education, 49(1), 7–24. https://doi.org/10.1080/03634520009379188