Situational speaking anxiety in adults who stutter

ABSTRACT

Purpose: Adults who stutter (AWS) are reported to have increased levels of anxiety compared to adults who do not stutter (AWNS), particularly in social interactions. However, the level of perceived anxiety according to specific speaking situations has not been critically evaluated. The present study examined the relationship between a physiological measure of anxiety (i.e., Cortisol) and perceptual judgments of communication apprehension across different speaking situations.

Method: Ten AWS and ten sex- and aged-matched AWNS provided salivary cortisol samples during four speaking situations across a one-week period. The speaking situations consisted of (1) speaking with a friend, (2) speaking with a single stranger, (3) speaking in front of a group of four strangers, and (4) speaking to a stranger on the telephone. Each participant also provided self-perception assessments of their perceived anxiety levels before and after each speaking situation.

Results: Significant group differences were found in self-perceived anxiety levels in speaking situations, however no such differences were apparent for the cortisol measures.

Conclusion: Overall results suggested a relationship between communication apprehension and various speaking situations among AWS, which is most evident in the self-perceived anxiety associated with speaking.

KEYWORDS:

• Anxiety
• cortisol
• self-perception
• environment
• stuttering

Acknowledgement

The research reported here was conducted by the first author and submitted in fulfilment for the requirements of the Master of Science degree at the University of Canterbury. The second author served as the primary supervisor.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 A power analysis was performed based on a sample size of 10 AWNS and AWS participants. The baseline results from the only two previous studies examining cortisol levels in AWS were considered as part of the analysis (Blood et al., 1994; Vasaghi-Gharamaleki, Mirzaii-Dizgah, & Arani-Kashani, 2016). The sample size of the Blood et al. (1994) study was 11 AWNS and AWS, while the sample size of the Vasaghi-Gharamaleki et al. (2016) study was 19 AWNS and AWS. The subsequent power analysis based on two-tailed criteria with 10 participants per group yielded a probability of 1.0 of correctly rejecting the null hypothesis when it is false.

2 As part of the post-speaking analysis, it was noted that one participant exhibited a cortsol value above the expected threshold (AWS8 at T-post). This value was confirmed by repeat analysis. The data point could not be ruled out for blood or another source of contamination of the saliva. Therefore, the statistical analyses of the AWS group post-speaking cortisol data were performed twice. The data were first analyzed with the data point from AWS8 (T-post) included and then analyzed a second time with the data removed and treated as an outlier. In both instances, there was no difference in the overall result. Therefore, the results of the statiscial analyses are reported with the data point from AWS8 included.