The non-significance of these relationships are hypotheses to be tested. Hence, the model testing analysis can be logically broken into two components. First, the statistical significance of those causal relationships hypothesized to be significant will be tested.