ilabelled was developed to address a fundamental problem in labeling values in R:
In addition, the i_labelled class should encourage users to think directly about meta information about variables in the work process and add it to the variables.
The basic idea behind this endeavor is to enable a similar handling of survey data as is the case in common social science statistical software such as SPSS or STATA.
R is a great software for working with data. However, there is a non-negligible disadvantage when it comes to handling categorical data and meta-information, such as variable labels or scale levels, and the like, as known from other statistical software, especially SPSS.
The aim of this package is to introduce a data class in which meta information (variable-label, value-labels, missing-values, scale-level) can and should be directly included. In addition, one of the biggest disadvantages of the base R factor class is to be addressed: Numerical values are not changed if they are provided with value labels. Negative and non-sequential consecutive values are not changed. In other words: the underlying database is not changed by labeling. Nevertheless, the option of using value labels for indexing is offered.
An intuitive syntax enables easy access to package-specific functions. All functions available for data manipulation can be controlled by an “i_*”.