Anyone conducting empirical research is familiar with the problem of missing data. Analysing incomplete data is complicated and time-consuming and can lead to erroneous conclusions. That is why we are working with TNO to develop new methods for analysing incomplete data.
As early as the 1990s, TNO created a computer algorithm designed to replace missing data with synthetic values, known as Multivariate Imputation by Chained Equations (MICE). We refer to this type of data replacement as imputation. Scientists use this MICE algorithm when they encounter missing data problems in genuine data. Utrecht University is currently investigating the statistical characteristics of this algorithm in order to improve it. Two dissertations and several research projects have already been dedicated to the subject. In the future, MICE will be transformed into a method for estimating individual causal effects. This will turn MICE into a reliable method for personalised advice.