Survey research problems during fieldwork with personal interviews in epidemiological studies (in Spanish)

Co-authored with Clara Llorens, Salvador Moncada and Albert Navarro. Published in Metodología de Encuestas, 2006, volume 8: 61-74 (Monográfico: Incidencias en el trabajo de campo).

OBJECTIVE: The aim of this article is to describe which are the most common incidences during data collection through personal surveys on epidemiological studies. METHOD: With this finality, a data base containing the reported incidences for wave 2’s «exposición a factores psicosociales en el trabajo, precariedad laboral, doble presencia y salud» study was constructed. Number of surveys for wave 2 was 2.540, the total number of waves was 3, and the total number of surveys was 7.650. This data base included, leaving aside the 7 dependent variables for each type of incidence, 3 variables serving to describe, from a sociodemographic point of view, the census blocks where the surveys took place. Data base was analyzed using Crosstabs and Chi-Square Test to obtain measures of association, because this technique, half-descriptive half-analytical, is adequate to investigate unclearly-defined objects of research (and this is the case here). RESULTS: Results suggest that must be considered a two-level-approach in description of incidences —the number of inhabitants in the municipality, as the first level, and sociodemographic features of the census blocks, as the second level— because incidences are differently related to census blocks’ characteristics depending on the number of inhabitants of the municipality: for example, in the provincial capital, number of immigrants are associated to certain incidence, while the same incidence are associated to another characteristic (for example, ageing of population of the census blocks) in a village.