Hi all,

I'm looking for some advice on data analysis. I can't see the wood for the trees now at this stage.

I'm doing a study of the effect of organisational context and Intranet system quality on knowledge sharing.
I have 9 independent variables and one dependent variable.
4 of the independent variables are various aspects of organisational climate: trust, innovation, communication and top management support.
3 are various aspects of organisational structure: formalisation, centralisation and rewards
2 comprise intranet system quality in terms of ease of use and perceived usefulness.
I have developed seperate hypotheses for each of these 9 items so they will be treated seperately.
These are my hypotheses:
  1. H1: An organisational climate high in trust will positively influence employee knowledge sharing
  2. H2: An organisational climate characterised by high levels of affiliation will positively infuence employee knowledge sharing
  3. H3: An organisational climate characterised by higher levels of centralisation will negatively affect employee knowledge sharing
  4. H4: An organisational climate characterised by high formalisation will negatively influence employee knowledge sharing
  5. H5: A positive communication climate will will positively influence employee knowledge sharing
  6. H6: Innovative organisational climates will positively affect employee knowledge sharing
  7. H7: An organisational climate where knowledge sharing is recognised and rewarded will positively influence employee knowledge sharing
  8. H8: Individual perception of top management support will postively influence employee knowledge sharing
  9. H9: High ease of use of information technology will positively affect knowledge sharing
  10. High perceived usefulness of information technology will positively affect knowledge sharing

All of the items will be collected using Likert scale data.

The research objectives I aim trying to answer with this data are:
1. To establish employee levels of knowledge sharing
2. To assess the impact of organisational climate on knowledge sharing
3. To assess the impact of organisational structure on knowledge sharing
4. To investigate the impact of information system quality on knowledge sharing behaviour

I understand that Likert data is ordinal and there seem to be varying opinions about whether this can be treated as continuous for the purpose of using regression analysis. The only method for examining relationships for ordinal data in Saunders (2007) is Spearman's. But this only considers two variables. My hypotheses only contain two variables. However it seems to me that I need to examine the effect of all these variables simultaneously as I will not know the true impact of the independent variables unless I analyse them all together. Am I right in saying this?

My understanding is that if the data is normally distributed and linear, I may use linear multiple regression. If not, I will have to use ordinal regression.

Am I on the right track here?