Hi, I have just read Ramaswamy 1993 as this is the source that everybody cites for the post hoc analyses.
I need some advice on how to bring this together with the data given in Smart PLS / FIMIX.
I am reffering to page 115 - 4.5 Diagnosing the pools - equation (27)
The equation looks straigt forward, I just need some help on the data that should be used. Here's what I understand:
0. I understand that pool k is the same as segment k
1. What would a SBU be in my case? one dataset, i.e filled out questionaire?
2. Z(ir) are the values for my indicators and demographic control variables, right?
3. Should I take the original values from the questionaires or some computed values from Smart-PLS?
4. delta(rk) is called the "impact ceofficient for variable r for pool k" => What is this in Smart-PLS? the FIMIX Final Partitions?
4.1 what value for v(ik)?
5. What do I need to compute to get Table 5? Because
5.1 Table 5 seems to have explanatory variable r per column and
5.2. not the SBU i as it is suggested in equation (27) for Q(ik)
5.3 how to standardize the coefficients as described in "results" on page 118?
6. and finally: how do I get the significance levels / t-values?
7. and Intercept
8. and Likelihood Ration X^2