The Rsocialdata package: Handling survey data in R

Population studies strongly rely on survey data and much time is needed to prepare the data. Assigning comprehensible short and long labels renders outcome more directly usable, and producing a detailed summary informing about the distribution of the variables is essential for efficiently documenting the collected data. The Rsocialdata package is intended to help the demographer or analyst in this task, allowing him to focus more quickly on the analysis. The toolbox come in the form of a series of R packages. It accepts user-defined missing values and then allows to easily turn a missing value as a valid case and vice-versa. It natively account for weights when available and process automatic checks to prevent the loss of representativeness when filtering out cases with missing values for example. As all information is stored within the data object a method for generating a codebook is provided. Furthermore, the toolbox provides efficient methods for handling panel data organized in successive waves. For example by specifying '..' in place of the two year digits in the variable names, the user can extract a whole sequence in a single step, recode some values, or turn a missing value into a valid case directly for all waves where the variable exists. In this paper we introduce some key functionnalities of our toolbox.