The January 2017 issue of the Journal of Voice (v31 n1) has an article by Anna Carolina Ferreira Marinho et al. titled Fear of Public Speaking: Perception of College Students and Correlates. You can read the abstract at PubMed, and a preprint of the full article at ResearchGate.
1135 students at one college or university replied to a questionnaire: 765 women (~2/3) and 370 men. As is shown above in a bar chart, 63.8% of them feared public speaking – 68.8% of females and 53.8% of males (significantly less). Age and field of study did not significantly affect the extent of fear.(Click on the chart to see a larger, clearer view).
A second bar chart shows how students perceived their own voice. 36.8% thought it was adequate. The majority did not – 30% thought it too high pitched, 18.5 % thought it too soft, 7.4% thought it too nasal, 4.9% thought it too deep, and 2.4% thought it too hoarse. An overwhelming majority of 89.3% expressed interest in getting speech language training.
The results shown in my first bar chart came from Table 1 of this article. If you go back and look at it you will find I calculated the percentages from the numbers rather than using those shown directly there. Those percentages are wrong – they were calculated by columns rather than rows. This is the sort of nonsense which can result in an article with multiple authors when a table is delegated to someone else but not carefully checked.
Data by sex from Table 1 are shown above. For males, the percent answering yes should be 100*(199/[171 + 199]) or 53.8%. Instead it was 100*(199/[199 + 525]) or 27.5%. Obviously that percent for males should NOT depend on the percent of females who answered yes, and the row sum for yes and no should be 100%.
Data from Table 1 by age and field of study are shown above in two more tables. For age the percentages again were calculated by columns rather than rows. For field of study, the column sums reveal that the numbers for no and yes were wrongly switched as well.