Minimum Sample Size Requirements for a Validation Study of the Birth Satisfaction Scale-Revised (BSS-R)

The 10-item Birth Satisfaction Scale-Revised (BSS-R) is a theoretically anchored and easy to administer multidimensional measure of the birth satisfaction construct. The use of the BSS-R Internationally has led to an increasing number of translation and validation studies being conducted. An important onsideration for any validation/translation study of the measure concerns sample size. However, sample size estimations for validation studies are invariably based on ‘rules of thumb’ that are insensitive to the dynamics of the measure under scrutiny and may consequently lead to underpowered investigations. The current study sought to determine empirically the minimum sample size for a validation study of the BSS-R. Methods: A Monte Carlo simulation study was conducted using the parameter specifications of the original BSS-R validation study as the input model. An extensive series of simulations were conducted to estimate statistical power and simulation quality for a range of sample sizes (N = 50 to N = 1000). Sample sizes from published BSS-R studies were also included in the simulations conducted. Results: Monte Carlo simulations revealed the minimum sample size for a validation study of the BSS-R to be N = 175. The original BSS-R development study and the US validation study were found to be adequately powered and satisfied all quality criteria for the simulations. Two published BSS-R studies had insufficient sample size to assure confidence in avoiding type 1 error. Conclusion: Sample size estimation for validation studies should be empirically informed to avoid type 1 error and ensure an adequately powered investigation.