At Psychology Today on August 31, 2022 there is a useful article by Amanda Nimon-Peters titled You’re nervous about public speaking. She covers a lot of ground. But she begins by claiming:
“Studies from around the world indicate that anxiety related to public speaking occurs in more than 75 percent of the adult population [Ref. 1, Ref. 2].”
Is that true? No, unfortunately it just is a myth. Her reference 1 is to another article by Erica Crome and Andrew Baillie in the Journal of Anxiety Disorders (Volume 28, 2014, number 5, pages 471 to 479) titled Mild to severe social fears: Ranking types of feared social situations using item response theory. Table 1 in that article presents results for biased subsamples (subgroups of people with fears) from four different surveys: two from the United States and two from Australia. The first one from the United States is the National Comorbidity Survey – Replication (aka NCS-R). The second one from the United States is the National Epidemiologic Survey in Alcohol and Related Conditions (aka NESARC). Those two Australian surveys are the 1997 and 2007 versions of the National Survey of Mental Health and Wellbeing (aka NSMHWB). Reference 2 is a web link to the description of the research methodology for the 2007 survey, and is redundant.
A table shown above lists the percentage for fear of public speaking reported for those four surveys in Table 1. It also shows the sampling for each survey. There is a total sample size, n, a response rate in percent, and the product of those two, the number n0 responding to the survey. That table also lists the number n1 for the biased subsample listed in the title for Table 1. Amanda was confused by, and misinterpreted the data presented in Table 1. A biased subsample is quite different from the random sample of size n0 that responded to the survey, which can be validly used to estimate a percentage for the population. Furthermore, in the 1997 NSMHWB fear of public speaking was at just 55.1% rather than more than 75%. (But when we average results from all four surveys, we get 76%).
For example, in the NCS-R n = 9282, and the response rate is 70.9%, so n0 = 6581. In Table 1 n1 is only 2261, so that biased subsample is just 34.4% of n0, as is shown above via a Venn diagram.
Percentages for various fears in the NCS-R from Table 1 are shown above via a horizontal bar chart. The largest, 87.9%, is for public speaking or stage fright (Acting, performing or giving a talk in front of an audience).
How does that compare with the percentages for the full random sample? Those results can be found in a 2008 article by A. M. Ruscio et al in Psychological Medicine (Volume 38, number 1, pages 15 to 28) titled Social Fear and Social Phobia in the United States: Results from the National Comorbidity Survey Replication, whose full text can be found here at PubMed Central. The first column in Table 1 of that article shows for NCS-R, the percentage for public speaking or stage fright is just 21.2%, or less than a fourth of that reported by Crome and Baillie. The last column shows results for fears among those (n2 = 1143) with lifetime social phobia, where 88.7% feared public speaking or stage fright. Those results are shown above via another horizontal bar chart.
Back on February 3, 2014 I had blogged about Busting a myth – that 75% of people in the world fear public speaking. Eight years later I’m again busting a version of that myth. There is a similar myth with 73% which I blogged about on March 22, 2019 in a post titled An apparently authoritative statistic about fear of public speaking that really lacks any support. I first blogged about results from Ruscio et al in a post way back on June 23, 2009 titled You are not alone: fear of public speaking affects one in five Americans.
Using the correct percentage (21.2% rather than 87.9%) from the NCS-R results in an average for fear of public speaking from those four surveys of 59.3%– a drop of 16.7%.
UPDATE
I emailed Amanda and received the following reply on September 13th:
Dear Richard,
No comments:
Post a Comment