The Right Analysis for the Right Data in Aesthetic Surgery Research

The article titled Did She or Didn’t She? Perceptions of Operative Status of Female Genitalia by Sasson, Sharp, and Placik is a cross-sectional survey of 511 adult participants and 21 aesthetic vulvar surgeons with the aim of evaluating lay individuals' and healthcare professionals' ability to identify participants who had undergone labiaplasty.1 Here, the authors conclude that both groups demonstrate difficulty identifying individuals who had undergone labiaplasty from images alone. Although we commend the authors for their work, our research team would like to comment on the measures of association and the isolated use of P values within the manuscript. First, the study authors utilized the Pearson correlation coefficient (r) to report the association between “natural” and “aesthetic” respondent ratings, measured using 2 ordinal Likert scales (ie, 1-5 scales). Although the use of Pearson's coefficient to measure associations between ordinal data is not uncommon within the published academic literature, it is important to note that this analysis should typically be reserved for 2 continuous variables that are normally distributed and demonstrate a linear relationship. For the comparison of 2 sets of ordinal data, the Spearman rank correlation coefficient (ρ) or Kendall's coefficient of rank correlation (τ) is recommended. These analyses do not carry the same assumptions about the distribution of the data and are calculated with the ranks, rather than the actual values, of the 2 variables. Further details of these tests as well as recommendations for their application are reported elsewhere.2,3 Second, throughout the manuscript the authors frequently employ P values as the sole evidence (or lack thereof) of an association between variables. Again, although this is not uncommon in the peer-reviewed literature, our research team wishes to highlight the concerns that stem from the reliance on P values in isolation to establish conclusions. Specifically, P values can be small even for trivial associations, which may not be clinically significant.4 Without reporting the actual correlation coefficients alongside the P values, readers cannot assess the strength and practical clinical significance of the associations, potentially leading to overinterpretation of results. The P value simply indicates the probability of obtaining the observed result, or a more extreme result, under the assumption of no effect (ie, null hypothesis) for a particular statistical test.4 Notably, the P value does not measure the strength or the importance of the result, but rather the size of the effect measure does. Especially in the context of sample sizes that are large, statistically significant P values may not correspond to meaningful or strong relationships.5 Ultimately, our research team feels that none of these criticisms pose a threat to the findings of this manuscript, and the authors correctly suggest that additional research is needed to confirm these novel and interesting results. Although photographic evidence continues to be important in evaluating aesthetic interventions, the utilization of validated tools may assist in conducting and interpreting data. We hope this manuscript serves as an opportunity to highlight these common statistical concerns as well as to elevate research in the specialty and in this journal. The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article. The authors received no financial support for the research, authorship, and publication of this article.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.