forum.smartpls.com

Posted: **Fri May 27, 2016 10:00 am**

I have a model where four of my variables come from entiment analysis of text. These are: C2B_PTP, C2B_PTS, C2B_SP, C2B_SS.
For some reason, they are showing me a very very high t-statistics (see figure). Also when running PLS Algorithm, their factor loadings are 1.000.
I already checked the data, and these variables are not proportional. They have correlation, but it's not perfect correlation. So I don't know what could be wrong here.
Anyone know what could be the reason my variables have 1.000 factor loading and huge t-statistic?

Posted: **Mon Aug 08, 2016 2:57 pm**

Maybe during the import there was a problem with the decimal sign so that you important numbers in a different format.

Posted: **Wed Aug 30, 2017 6:57 pm**

Same problem is with me.. my one variable factor loading value is 1 (as pasted below)... what is the reason?

: FL-1.jpg (79.44 KiB) Viewed 62256 times

: FL-2.jpg (70.99 KiB) Viewed 62256 times

Posted: **Thu Aug 31, 2017 7:08 pm**

It is a single-indicator construct. Those have per definition in PLS an outer loading of 1, i.e., construct and indicator are per definition equal.

forum.smartpls.com

Huge t-statistic. How come?

Huge t-statistic. How come?

Re: Huge t-statistic. How come?

Re: Huge t-statistic. How come?

Re: Huge t-statistic. How come?