forum.smartpls.com

Hi Bido, Thank you for your reply. All the models I work with will have more than two LV. usually between 5-12. However, sometimes it can be difficult to determine the relationship. E.g consider these three variable I (image), P(product), R(Re purchase) My model would then be (I-->R) (P-->R) However...

Dear all, Is it possible in smarpls, to determine whether a relationship goes from A to B, or B to A? Most often I naturally see the same correlation when reversing the arrow between to LV. However, is it possible, without considering the logical and chronological order, to mathematical determine in...

Hi Bido, Thank you for the response and reference. The article raises some surprising and concerning questions. It would seem I should use the mean replacement, and then just accept the lower R^2 and path coefficients, despite my belief than casewise deletion gives a more "clean" model. Ca...

Dear all, Has anybody any experience in using both casewise replacement and mean replacement, in the same model. My problem: 1) if using mean replacement, one of my LV's are not significant, this is due to too many missing values, which is a bias to the variance, however, all empiric studies and sim...

Dear professor, A have a PLS model with 6 LV LV1 har a path towards LV2, LV3, LV4, LV5 and LV6 My problem is, that the path from LV1 towards LV2, is high, but has a very low R^2. (path = 0,271 : R^2 = 0,074) Does this emply that the path coefficient is not reliable? Is it still "safe" to s...

Hi Bido, A question to your answer above: 3) In the report for validating formative construct which t- values should i see. ( outer weights or outer loadings t values) Outer weights Does this mean, that we in a formative LV can remove indicator variables rejected in the t-test? Are this then interpr...

Hi Christian, Regarding the t-test on the outer weights in a formative model, can these be used to test whether the indicator provides significanse support to its LV? I seem to have read in much litterature that the only way to test this in a formative model, is by a vanishing tetrad. BR Hello, I am...

Hi professor, If one of the outer weights in a formative model is negative, is this then equal to that this indicator has negative correlation with the dependent variable? or how should such a negative weight be interpreted. When just doing a linear regression with this one indicator and the indicat...

Hi Bido, I have a LV containing 5 manifest variables, when doing a principal component factor analysis in spss, all variables load with more than 0,5. Hence I am concluding that no variables needs to be excluded. However, when entering the model in smarpls, the t-values for 2 of the 5 manifest varia...

Hi Bido, Thank you very much for the fast reply. I know how 3 different follow-up questions :) Do you have a reference to a research paper or similar that explains the different weighting schemes in smartPLS? (factor, path, centroid) Do you know if there is any advantage in "cleaning" your...

Hi Howards,
Thanks, look forward to reading it

BR

Hi Diogenes, Thank you for the reply. Yes the path coefficient are as I understand the standardized versions of the linear regression weight results. When describing the relationship between the LV, and calculating the impact on one LV when increasing another, which measurement would be recommended ...

Hi Professor, I have a question regarding the t-testing between the LV. If SmartPLS dosnt do this, then what are the results shown in "Path Coefficients (Mean, STDEV, T-Values)" when doing a bootstrap? You get these results "T Statistics (|O/STERR|)" which I thought were the t-te...

Dear professors, I want to make a model which measures the impact on loyalty towards a brand. I have a formative model with 4 latent variables, "image", "product", "service" and "loyalty". Each of these concist of several measurement variables. Image, Product ...

Dear Bido, I have a total of 409 respondents in my data, when bootstrapping, I select 409 cases and 1000 sample. When I select to use all my cases in each sample, how come I dont get the same result with 1000 samples, as I would get with 100 samples? (as all my cases are used in each of the 1000/100...