Abstract
The international assessment studies of key competences, such as the PISA report of the OECD, have revealed that the academic performance of Spanish students is significantly below the OECD average. In addition, it has also been confirmed that the results of immigrant students are consistently lower than those of their native counterparts. Given the context, the first objective of this work is to observe the variables (support, control, school satisfaction, and learning environment) which distinguish between retained and non-retained native and immigrant students. The second objective is to check, by comparing the retained and non-retained native and immigrant students and separating the two levels, in order to find out which of the selected variables clearly differentiate the two groups. A sample of 1359 students was used (79.8% native students and 20.2% immigrant students of Latin American origin), who were enrolled in the 5th and 6th year of Primary Education (aged 10-11 years) and in the 1st and 2nd year of Secondary Education (aged 12-13 years). The measurement scales, which undergo a psychometric analysis in the current work, have been developed in a previous research study (Lorenzo et al., 2009). The construct validity and reliability are reported (obtaining alpha indices between 0.705 and 0.787). Subsequently, and depending on the results of this analysis, inferential analyses are performed, using as independent variables the ethno-cultural origin and being retained or not, whereas, as dependent variables, the indices referring to students' perception of family support and control, as well as the assessment of the school and learning environment. Among other results, the Group × Being retained/Not being retained [F((1, 1315)) = 4.67, p < 0.01] interaction should be pointed out, indicating that native non-retained subjects perceive more control than immigrants, as well as the Group × Being retained/Not being retained [F((1, 1200)) = 5.49, p < 0.01] interaction, showing that native non-retained students perceive more family support. Given the results obtained, our intention is to provide solid evidence that would facilitate the design of family involvement programs, helping to improve students' educational performance.