Hi Pooneh,

we usually do this by hand and not automatically, e.g., we would build groups from the non-likert scale questions, e.g., by age and compare the likert-scale results for each group with other groups. Typical results would look something like: in the age group of above 50, the average evaluation for the importance of the quality of life is at 4,5 in contrast to the age group below 30 where it calculates to 2,3. 

Afterwards you should also check the significance (e.g., by using a simple t-test). For the questionaire, we usually use Cronbach's Alpha in order to check the internal consistency of the questionaire.

Hope this helps.

Cheeers,

  Marc

You can compute a product-moment correlation between your Likert scales and the dichotomous items. For the categorical variables that have more than two categories (e.g., where someone lived), you could do analysis of variance with the Likert scales as the dependent variable. You can compute effect size--there is more than one way to do that.

Hi Pooneh ! 

Hope I got your name right?  Well data in statistics are broadly categorised into two,  categorical and continuous.  The type of analysis you can do depend on the data type. Likert scale responses are score and assumed to be a continuous measure,  in that case,  pearson correlation can be used to measure the magnitude and direction of relationship between two variables ( two or more questions in your case)  and if you desire a more robust analysis,  you can do a pairwise comparison using the test.  Hope this is helpful?  Best of luck

Pooneh,

I think the best way to achieve what you want was pointed by Paul E. Spector; for example, a logistic regression of your lickert variable on your categorical variables could give you effect sizes.

Cheers!