Following the practice of the SEM model development approach
(Kline, 2005), alternative models were tested. Model 1 included seven
exogenous variables (i.e., gender, age, extraversion, agreeableness,
conscientiousness, neuroticism, and openness) that were not predicted
by other variables in the model. Model 2 replaced age with level of
study, a closely associated variable. For comparison purposes, an extra
model (Model 3) was computed to include only the statistically
significant paths. In all three models, SNS use was conceptualized as
the mediating endogenous variable. The perceived usefulness of the
acquired everyday life information in meeting everyday needs was
the final endogenous variable (see Fig. 2).