[urgernt]Multi-Cultural, and job tenure influences

Research topics can be discussed in this area.
Post Reply
willywonka78
PLS Junior User
Posts: 4
Joined: Fri Mar 03, 2017 1:48 pm
Real name and title: Chu, MBA student

[urgernt]Multi-Cultural, and job tenure influences

Post by willywonka78 »

Hi, I would like to address to the same question here. I am very new to SmartPLS but thanks to the material on internet I am able to understand how to use it little by little.

I am doing my master thesis about what factor will influence employee to use company's internal social media, and the hypothesis is that based on different cultures and level of job tenure, and goal of using it will influence the usage.
So I have around 125 responses from survey based on 6 likert scales,
here is the question, I want to do the analysis based on different region (Europe, North America, South America, Asia, Africa and so on) and also different job tenure(1-5, 6-10years, 10 years above, and also different goal (work related goal or networking related goal)

my question is, how do I use the Multi Group Analysis to separate different regions, job tenure, and goals at the same time?
it is possible to do that? oh yes and I am using the student account so I think I am not able to use more than 100 responses.


thank you for answering me!
User avatar
cringle
SmartPLS Developer
Posts: 818
Joined: Tue Sep 20, 2005 9:13 am
Real name and title: Prof. Dr. Christian M. Ringle
Location: Hamburg (Germany)
Contact:

Re: [urgernt]Multi-Cultural, and job tenure influences

Post by cringle »

Thanks. You can do a PLS multigroup analysis using SmartPLS 3. Take a look at the PLS-MGA videos: https://www.youtube.com/channel/UCW4GA7 ... sCgSDLuo1Q

However, you most likely need more the 100 observations. Get SmartPLS 3 Professional or use the 30 days free trial (see FAQ on www.smartpls.com on how to get it).

Best
Christian
willywonka78
PLS Junior User
Posts: 4
Joined: Fri Mar 03, 2017 1:48 pm
Real name and title: Chu, MBA student

Re: [urgernt]Multi-Cultural, and job tenure influences

Post by willywonka78 »

Dear Prof. Dr. Ringle,

let's say that having more responses is not possible anymore....I am afraid that this is the limitation of this research

today I tried to run a MGA based on a group of all south American and other region, i did exactly the same as it was told on the video.
I was afraid that it might show me error, because there are only 10 response from South America and therefore Group A has only 10 and Group B has 90
and it did.

the error message is this one below, does it mean that I have too less sample for Group A(south america)?


Thanks for your reply!

Singular Matrix Problem

A singular matrix is a square matrix that does not have a matrix inverse. This is a problem in the calculation of regression coefficients and correlations where the inverse of the matrix plays an important role.
Problem

There can be several reasons that such a singular matrix problem occurs during the calculations.Possible reasons could be
a variable that has zero variance, i.e. it contains only the same values for each case/observation. This can happen, for example, if one uses a grouping variable in the path model and estimates the group models.
extreme collinearity, for example, if two variables are perfectly collinear (i.e., they have a correlation of 1).
a sample size that is too small. If there are fewer observations than degrees of freedom the matrix of moments is not invertable.
How to Solve It?

Identify variables that may cause the problem and remove them from the model or increase the sample size.

cringle wrote:Thanks. You can do a PLS multigroup analysis using SmartPLS 3. Take a look at the PLS-MGA videos: https://www.youtube.com/channel/UCW4GA7 ... sCgSDLuo1Q

However, you most likely need more the 100 observations. Get SmartPLS 3 Professional or use the 30 days free trial (see FAQ on http://www.smartpls.com on how to get it).

Best
Christian
User avatar
cringle
SmartPLS Developer
Posts: 818
Joined: Tue Sep 20, 2005 9:13 am
Real name and title: Prof. Dr. Christian M. Ringle
Location: Hamburg (Germany)
Contact:

Re: [urgernt]Multi-Cultural, and job tenure influences

Post by cringle »

Well, 10 observations is way too small for a model estimation.

The "singular matrix" error message usually results from one variable in your model that only has identical values. Then, there is no variance and the algorithm does not work. This issue can occur for the full set of data or per group.

Best regards
Christian
Post Reply