ABSTRACT
Psychological research frequently encounters criticism regarding the representativeness of the samples under study, highlighting concerns about the external validity of the obtained results. Here, we conducted a comprehensive survey of all the quantitative samples from the journal Body Image for 2021 (n = 149 samples). Our primary objective was to examine the extent to which the sampled populations deviated from the population at large, which could potentially compromise the generalizability of findings. We identified that a substantial number of these samples came from student populations (n = 44) and the majority were from the United States, United Kingdom, and Australia. Only a small number of samples (n = 9) employed direct measurements of body mass index (BMI), while the majority relied on self-reported data (n = 93). For a subset of samples in the journal, which were drawn from the general population, we compared whether these differed from population reference values in terms of age and BMI. Using Monte Carlo simulations, we found that samples tended to be younger and score lower on BMI than reference values obtained from the broader population. Samples drawn from female university students also tended to be lower on BMI than age-matched reference samples. We discuss the implications of our findings and make recommendations on sampling and inference. We conclude that a clearer specification of the parameters or conditions under which findings are expected to generalise has the potential to enhance the overall rigor and validity of this field of research.
