Replication files for "Analyzing Large-Scale Freediving Competition Data With a Special Focus on Gender Differences"

Abstract

Large multinational associations such as the International Association for the Development of Apnea (AIDA) or the Confédération Mondiale des Activités Subaquatiques (CMAS) hold freediving competitions regularly and provide the results to the general public. Their databases are a highly reliable source for analyzing overall freediving results in the three most relevant disciplines (static breathholding, STA; distance freediving, DYN; deep freediving, CWT). For the first time, we analyze these large-scale datasets (N > 11,000; share of women = 33%) and report a wide range of descriptive results such as median outcomes based on personal bests: STA 256sec; DYN 102m; CWT 46.5m. The correlations between the three disciplines are between 0.52 and 0.65. Regression analyses reveal that men have better freediving results than women, on average, with an advantage of between 7.5% and 10%, depending on the discipline. Quantile regression models demonstrate that the gender gap grows with overall performance. It is the largest among elite freedivers but smaller among low- or average-performance freedivers.

Citation Bittmann, Felix (2025). Analyzing Large-Scale Freediving Competition Data With a Special Focus on Gender Differences, in: Sports Research International. https://doi.org/10.1556/1020.2025.00014

Data sources and imputation Data are taken from the official websites (AIDA / CMAS). The data have been imputed using MICE with 25 datasets. The final number of N depends on the specific disciplines, see the descriptive table below.

Descriptive results

STA by quantile and gender Explanation: if a male freediver has a static time of 305 seconds, he has a better competition result than 70% of the male competition sample.

DYN by quantile and gender

CWT by quantile and gender

Association of disciplines

Association of STA and DYN by gender

Association of STA and CWT by gender

Association of DYN and CWT by gender

Descriptive table

Correlation table

Note: Pearson's r. Pairwise case numbers in parentheses. All correlations are statistically significant on the 0.1% level.

Gender differences (Regression table)

Data

Due to data protection issues, we have encrypted the datafile ("data.7zip"). You can get access by writing an email to the author (felix.bittmann@lifbi.de or felix.bittmann@uni-bamberg.de). If possible, use your institutional email for contact. Use the free software 7-Zip to extract the archive (https://7-zip.org/).