Multiple imputation with chained equations (MICE) is a widespread approach to account for missing data in empirical research. Combining MICE with bootstrapping, that is, repeatedly resampling with replacement from the data to estimate variances and confidence intervals of statistics of interest, is not straightforward. The current document provides an overview of how to use bootstrapping with imputed data in Stata. Two main approaches (impute first and then bootstrap or vice-versa) are discussed and shortly compared. Code is provided for Stata.

For a long time, there has been interest in the relationship between personality traits and life satisfaction. However, many previous studies suffer from small sample sizes or selective sampling, sometimes leading to biased or imprecise results. The current study, which focuses on the trait of extraversion, attempts to overcome these limitations by using a large-scale German survey of more than 40,000 individuals aged 10 to 72. As the descriptive and regression results clearly show, there is a statistically significant relationship between extraversion and several dimensions of current life satisfaction. Depending on the type of satisfaction, the unstandardized regression coefficients range from 0.18 to 0.33. To gain further insight, age is added as an explanatory variable to test whether these associations change as a person ages. The results show that the effect sizes sometimes differ by age but remain robust and statistically significant. In conclusion, the study adds to the literature by providing convincing evidence that extraversion and life satisfaction are positively associated, at least in Western and industrialized countries, and that there are interaction effects with age.

Bootstrapping is a flexible, powerful and well-established statistical approach to quantify the uncertainty of virtually any point estimate. While multiple versions of bootstrap confidence intervals are already available in Stata, dbs implements the double (iterated) bootstrap. Instead of relying on parametric assumptions such as the non-parametric resampling bootstrap confidence interval does, it is more flexible and derives critical values directly from that data. To do so, multiple methods are available (analytic approach, double resampling, jackknife estimation). In a comparative simulation study it is empirically demonstrated that the strengths of the double bootstrap are particularly evident for small samples (n < 100) when heteroscedasticity is present. While all other approaches result in undercoverage, only the double bootstrap reaches the target coverage level and hence avoids incorrect statistical conclusions. The computational burden is not even necessarily larger than for other bootstrap approaches.