Abstract

A relevant research needs adequate methods and techniques of data analysis. With the aim of assessing criteria of relevance in the study of change, spontaneous or following specific events or treatments, the main techniques suitable for longitudinal data have been reviewed. After examining the problems posed by traditional statistical based on pre-post differences, some techniques available for analyzing time series are presented, including the search for critical change points; sequential and survival analyses, linear hierarchical or multilevel models. Categorical data typical of qualitative researches are also taken into account. Some considerations about differences between statistical significance and meaningfulness are pointed out, stressing practical relevance of the inferences which can be drawn from adequately analyzed data.