Speaking Fear and Dropping Out: The Four-Year Study of Who Left University
By Noam Shemla · The Studies
TL;DR McCroskey, Booth-Butterfield and Payne measured how anxious every new student at West Virginia University was about speaking, then followed all 1,884 of them for four years. The most apprehensive students were more likely to leave the university, and in their first two years they earned lower grades. By the third year, the students who were still there looked alike.
- Paper: McCroskey, J. C., Booth-Butterfield, S., & Payne, S. K. (1989). The impact of communication apprehension on college student retention and success. Communication Quarterly, 37(2), 100–107.
- Participants: 1,884 new students who entered West Virginia University in autumn 1982 and attended orientation; the sample matched the whole intake by sex and by dropout rate.
- Measure: The 24-item Personal Report of Communication Apprehension, taken at orientation. High meant a score above 80 (335 students by the text, 352 by the table), low below 51, the rest moderate.
- Outcomes: Official grade point average and enrolment, every semester for eight semesters. Any student without a grade in a semester counted as a dropout.
- Replication: The 1983 intake, followed the same way; only about half of it was measured.
- Access: Subscription-only at the journal. The first author posted a scan on his publications website, which is what this summary is read from.
Getting into university is one test; staying is another. James McCroskey had already shown that students afraid of speaking earn lower grades. With Steven Booth-Butterfield and Steven Payne he asked the harder question: whether that fear decides who is still there four years later.1
Why speaking fear might push students out
Research on who finishes university points to two kinds of success: academic, meaning the ability and study habits a student brings, and interpersonal, meaning the social ties they build once there. The authors argued that apprehension about speaking undermines both.
There was one earlier hint. A conference paper by Mehrley (1984) had found that, in their first semester at Indiana State University, 15.5% of highly apprehensive students dropped out against 8.4% of those with low apprehension. That was one semester at one university, and it said nothing about grades.
Following a whole intake
At the orientation every new student attends, the team gave the 1982 intake a standard 24-item questionnaire on speaking anxiety. For the next eight semesters they collected each student’s official grade point average and whether they were still enrolled. The rule for dropping out was strict: any student without a grade in any semester counted as a dropout, including those who transferred elsewhere or left for family reasons. The authors reasoned that this would blur the result rather than create one, since there was no reason those exits should depend on apprehension.
The sample looked like the whole intake: its two-year dropout rate was 29.5%, against 29.4% for all new students that year.
Who left
Students who left, by year and apprehension (first cohort)
| Low (335) | Moderate (1,197) | High (352) | |
|---|---|---|---|
| Year 1 | 9.6% | 10.7% | 12.5% |
| Year 2 | 14.3% | 19.4% | 20.2% |
| Year 3 | 8.1% | 8.2% | 8.2% |
| Year 4 | 3.0% | 5.2% | 5.4% |
| By year 2 | 23.9% | 30.1% | 32.7% |
| By year 4 | 34.9% | 43.4% | 46.3% |
Apprehension made the biggest difference early. The authors read this as a sign that the anxious students who survived the first two years had either arrived with ways of coping or learned them fast, or perhaps steered clear of courses that demanded a lot of talking. Either way, they concluded, help has to come in the first year.
Checking the arithmetic
The paper tests each difference between high and low apprehensives with a z-test for two proportions, and states its size as Cohen’s h. Recomputing both from the paper’s own table gives the same direction throughout, but smaller test statistics than the paper prints.
The four-year difference survives comfortably: p = .002 by the standard test. What does not survive is the first-year difference taken on its own, which the paper reports as significant (z = 1.71) but which the standard test puts at z = 1.23, well within chance.
Grades
In the first year, highly apprehensive students averaged a GPA of 2.58 against 2.76 for low apprehensives, a small difference (d = .23). In the second year the gap narrowed to 2.74 against 2.84 (d = .13), and in the third and fourth years it was no longer significant.
The authors then noticed something in their own data that is easy to miss: the average GPA of the whole group rose every year, for every level of apprehension. It was not that students were improving. When they looked only at the students who stayed all four years, grades changed little, and the difference between high and low apprehensives was not significant.
Among students who left, the pattern held too. In the first semester, highly apprehensive dropouts averaged 2.17 against 2.36 for low-apprehension dropouts, and in the second year 2.32 against 2.57.
What the authors made of it
The effect is small by any statistical measure, the authors stressed, and apprehension should not be expected to explain much of how a whole student body fares. In practical terms it is not small: in this intake, bringing the most anxious students’ dropout rate down to that of the least anxious would have meant roughly 40 more students finishing their degree.
They proposed three responses: treatment programmes to reduce apprehension, which they described as cheap; counsellors who steer anxious students away from the most talk-heavy courses and keep trying to reach students who avoid them; and teaching that does not grade on participation, offers alternatives to required oral work, and stops reading quietness as a lack of interest or ability.
Has it held up?
The dropout finding has. Ericson and Gardner followed two further groups of students over time and found, again, that highly apprehensive students were significantly more likely to drop out. They found no significant difference in grades, in line with this study’s finding that the grade gap faded after the second year.2
The link to grades is small but consistent across studies: a meta-analysis of 30 studies found higher apprehension reliably, if weakly, linked to lower cognitive performance.3
What this study cannot tell you
- It shows who left, not why. Apprehension was measured and dropping out was observed; the academic and social routes in between were argued, not tested.
- Some of its significance tests are too generous. Recomputed from its own tables, the four-year result holds, but the first-year difference and the second cohort’s differences are not significant on a standard test.
- The second cohort was compromised. Questionnaires went missing halfway through orientation, and the students measured were noticeably less apprehensive than the year before, probably because anxious students came to the later sessions.
- Dropping out was defined broadly. Transfers, illness and family reasons all counted, which the authors argue can only have weakened the result, but it also means “left this university” rather than “gave up on a degree”.
- One large university, in the 1980s, with few students from minorities and no data on race or age. The authors themselves caution against applying it to small colleges.
What a speaker can take from it
The cost of speaking fear is easiest to see at the moment it is paid: the question not asked, the office hour skipped. This study shows the bill adding up somewhere nobody would think to look, in who is still enrolled two years later, and it shows it arriving early.
That timing is the practical lesson. The first months in any new setting, a university, a job, a team, are when being able to speak up matters most, because that is when people build the ties that make them stay. Working on speaking fear before those months, rather than after, is working on it when it pays the most.
This article summarises a published paper, read from the scan its first author posted. The counts, percentages, grade averages and effect sizes in its text are the authors’ own, from Tables 1 to 3 and the notes. The recomputed z-test, the risk ratios, their intervals and the pooled estimate are ours, calculated from the paper’s tables.
Measured
- Speaking apprehension, by questionnaire at orientation
- Enrolment and grade point average every semester for four years, from university records
Inferred, not measured
- That apprehension causes leaving: the study is observational, and the routes through academic and social life were not measured
- That the second cohort replicates the first: it points the same way, but it was only half measured and its difference is not significant on a standard test
References
- McCroskey, J. C., Booth-Butterfield, S., & Payne, S. K. (1989). The impact of communication apprehension on college student retention and success. Communication Quarterly, 37(2), 100–107. https://doi.org/10.1080/01463378909385531
- Ericson, P. M., & Gardner, J. W. (1992). Two longitudinal studies of communication apprehension and its effects on college students’ success. Communication Quarterly, 40(2), 127–137. https://doi.org/10.1080/01463379209369828
- Bourhis, J., & Allen, M. (1992). Meta-analysis of the relationship between communication apprehension and cognitive performance. Communication Education, 41(1), 68–76. https://doi.org/10.1080/03634529209378871