Surveys are used to infer the level of social integration of immigrants. Item response theory helps to describe the relationship between responses to test items and latent traits of interest. However, in the presence of nonignorable missing data, which are omitted responses depending on the latent traits to be measured, estimates of the model parameters are biased. To account for nonignorable missing data, the quantity and quality of contacts between immigrants and natives (so called “social integration”) are taken into account through a linear function of the response propensity. Higher education, no intention to migrate again, young age, Albanian nationality , and declaring a non muslim religion or none, comparatively favor social integration.
|Titolo:||Dealing with Omitted Answers in a Survey on Social Integration of Immigrants in Italy|
|Data di pubblicazione:||2017|
|Appare nelle tipologie:||1.1 Articolo in rivista|