Hi,
I am researching whether employees' perceptions of internal communication have a significantly significant relationship or positive association with level of employee engagement and type of employee communication behaviour.
I believe that internal communication and employee engagement are ordinal variables, whereas employee communication behaviour is a categorical variable but I'm not certain.
My question is how would I go about analysing the Likert-scale questionnaire data for this in terms of determining statistical significance and correlation?
Thank you
You would have to describe your data in much greater detail before anyone could hope to answer your question. Best wishes, David Booth
Internal communication is measured using 5 dimensions, with items scored using a 6-point Likert scale, ranging from strongly agree to strongly disagree.
The employee engagement items are made up of 3 dimensions and are scored using a 7-point Likert scale items ranging from never to always. E.g. "At work I feel bursting with energy"
The employee communication behaviour items are split into 2 dimensions (one of which is made up of 2 sub categories) and are scored using a 6-point Likert scale ranging from never to always.
I hope this is a bit more clear. This is my first time doing such a thing and I am really stuck
As best I can tell the idea is to have regression models of the form DV=IV1 +IV2 +.... Employee engagement has 3 7point scales etc. same idea with communication behavior. If I have this right i would suggest ordinal logistic regressions where the DVs are the Likert scale values for 1 seven point scale from employee engagement. The IVs are Likert scales from internal communication. You are going to have a lot of regressions. I am assuming that the DV scales are independent of each other. So set up your models and your data sets and run each model with ordinal logistic regression. See the attached course notes for how to do that. If this all is not clear, keep asking. But if I understand you correctly, this should work. Best wishes, David Booth
This is a case of Test of Association that u can employ successfully.
You can use the primitive or crude method to quickly estimate the significance of a correlation coefficient (see Best and Kahn's Research in Education, 2006). You can also measure the determination of correlation coefficient - the squaring of r. This will show how much of the correlation between the variables is common variance.
Thank you all for your help. I think I've now gotten my head around correlation coefficient.
However, I am unsure whether since my IV consists of 3 dimensions that assess total internal communication effectiveness, if I need to find the mean of the variables before doing the ordinal logistic regression test?
And with the communication behaviour DV, as it is categorical is there a different test used for this? Communication behaviour is grouped into 2 categories - megaphoning (positive megaphoning or negative megaphoning) and scouting.
So in essence I want to see if:
1. internal communication that is perceived as effective results in high employee engagement
2. internal communication that is perceived as effective results in positive megaphoning and scouting
3. internal communication that is perceived as ineffective results in negative megaphoning
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS