Dear Smart-PLS community,
I've heard that negative signs can occur due to:
a. initialization of the regression analysis
b. coding errors
c. wording errors
d. interview situation
e. insignificant indicator values (0.1 or 0.25)
Can anyone tell me:
1) where I can check if the negative indicator occured during the initialization and how to prevent it.
2) if e. means that the value (weight or loading) or the significance of the indicator has to be compared with the threshold?
3) if e. means that indicators with a value (weight or loading) or significance (see 2) between -0.25 and 0.25 can be the reason for the negative path-coefficient?
Do you know any other reasons for negative signs?
Thank you very much,
Timo