The State and its citizens generate lots of data. Once stored and processed, data can help resolve questions in Social Sciences, where it is common to need data in a different level of aggregation than the data is presented. In election results, we would like to understand the decision of each voter or group of voters, but the data is only available at the polling-place level. This problem is known as ecological inference. In this work we compare two methods to estimate the share of votes for each party according to different age groups. This work is complemented with the analysis of votes between the first and second rounds, as well as a referendum during the first. This analysis was performed on the 2014 uruguayan presidential elections.

Comment: 6 pages, in Spanish, 3 figures