How do we evaluate the t-value found after bootstrapping of smartpls? Comparing it with one-tailed or two-tailed critical t-values?
For example, H1 hypothesis states that A is positively related to B, H2 hypothesis states that C is related to D.
Do we use two-tailed critical t-values for both H1 and H2?
or one tailed for H1 and two-tailed for H2?
Many thanks,
Burcu