Table 5 Stages of the methodology.

PART 1. Qualitative part to define EMEA and to measure the level of engagement.

First stage: EMEA is a compound of three dimensions as previously explained: information, interactivity, and action. According to these three dimensions a data sheet was made up to facilitate analysis. The EMEA data sheet was applied to evaluate different degrees of engagement that could be achieved through the sample’s websites.

PART 2. Quantitative part to clarify the relationship of the level of engagement (measured in the part 1) and other variables from the CCEW (2020): to test our hypothesis.

Second stage: ANOVA and Kruskal–Wallis tests were used to identify any significant differences between the EMEA stages in the sample for the quantitative variables (“Results” (Last recorded income and Last recorded expenditure), “Number of groups targeted”, “Number of ways to help” and “Number of things done”).

Third stage: structural equation methods based on variance and specifically partial least squares (PLS) via the SmartPLS 3.3.5 program (Ringle et al., 2015) was used to propose a theoretical research model (based on the previous results) to test EMEA and hypotheses 1 and 2. Justification for the choice of PLS was: (1) The objective was to identify potential relationships between variables.

(2) PLS does not impose any specific distribution assumption (eg. normality) for the indicators and does not need the observations to be independent of each other (Chin, 2010).

(3) PLS can estimate structural models with small samples (Chin & Newsted, 1999; Hair et al., 2019; Rigdon, 2016).

(4) PLS can estimate models with estimated measurements in Mode A (traditionally reflective) and in Mode B (traditionally formative without any identification problem) (Chin, 2010). Our model presents variables measured as compounds estimated in Mode A, which are those created from diverse elements that are generally correlated. In our research model there is a higher order construct (EMEA) with which aggregate “scores” of the compounds are used in subsequent analyzes (Chin, 2010).