iris_afandiphd
PLS Senior User
Posts: 28
Joined: Thu Jun 24, 2010 5:38 am
Real name and title:

Hi Everyone,

I would like to assess the structural model via path coefficients. The bootstrap re-sampling technique (Chin,1998) was used by 500 iterations of the original 248 samples.

After finished the BT analysis, I got the HTML Report as below:

My Question is, Which one of the report in that table of content suitable for checking path coefficient significance?. Do I need to calculate t-statistics = path coefficient / S.E?, for testing significance at p<0.05,p<0.01 or p<0.001.

which column shows the path coefficient
Thanks

Diogenes
PLS Super-Expert
Posts: 905
Joined: Sat Oct 15, 2005 5:13 pm
Real name and title:
Location: São Paulo - BRAZIL
Contact:
Hi Brisbane,

You should notice that the table “Path coefficients (Mean, STDEV, t-Values)” has these columns: Original Sample (O), Sample Mean (M), Standard Deviation (STDEV), Standard Error (STERR), T Statistics (|O/STERR|).

The first column has the path coefficients that were computed with all sample, they are the same values when you ran the PLS algorithm.

The last column are the t-values (t > 1.96 are significant at 5%), these values are showed in the model too.

The others columns are intermediate calculations to have the t-value in the last column.

Best regards,

Bido

iris_afandiphd
PLS Senior User
Posts: 28
Joined: Thu Jun 24, 2010 5:38 am
Real name and title:
Dear Prof. Bido,

Yes, just realized that the t-value calculated as the same t statistics from Z population, according to basic statistics.

t= sample/S.E

How to calculate p-value? from coeff. table?

How about degree of freedom?, I found it via this URL
http://www.danielsoper.com/statcalc/calc08.aspx

Thanks

Diogenes
PLS Super-Expert
Posts: 905
Joined: Sat Oct 15, 2005 5:13 pm
Real name and title:
Location: São Paulo - BRAZIL
Contact:
Hi,
d.f. = n - 1 (n = number of resamples in the bootstrap)

you could use the p-value calculator that you had quoted

or Excel --> see viewtopic.php?p=3797&highlight=#3797

Best regards,

Bido