Methodology for the reconstruction of citizen surveys in different time periods

In the evaluation of the impact generated by public policies, it is common practice to collect data in surveys carried out at different points in time. The study presents a comparison of two main methodological problems: (i) that questions asked are not always identical, and (ii) that data may not always be interpreted with the same methodology. To assess the extent to which the factorial structure is maintained, this paper proposes an original application of the principle of parsimony, like the well-known elbow rule in sedimentation plotting. The statistical contribution is validated through two publications of the Spanish Sociological Research Centre (CIS), published in different years. The results of the analysis show that, despite the variations mentioned above, the factorial structure of the responses remains the same, suggesting that the changes may be attributable to policy intervention.