Dear

Katharina Franke

The question is not clear.

is PDI is a regression model? or what?

Dear Katharina Franke,

35 and 25, and other weights for other Hofstede dimensions, are weights that were set by Hofstede so that the scores in the whole set of countries would be distributed between 0 and 100. As best I can tell, these weights are arbitrary, or perhaps more precisely they reflect Hofstede's judgment and "feel" for the data. If you accept this, then there is no need to adjust these weights, and indeed changing them would render your data inconsistent with common measures (impeding replication, e.g.). On the other hand, there is an extra term in the formula for each dimension, which you can control:

PDI = 35(m07–m02) + 25(m20–m23) + C(pd)

C(pd) is a constant that can be used to adjust all the values in your sample, if you want to set the minimum to zero. This does not affect the comparison between countries. Hofstede argues that C(pd) can be used to shift the value so they are in the 0-100 range. To be more precise, you can set the minimum or maximum this way, but this will not ensure a 0-100 range; that would require multiplying the scores if you have a set of countries with limited range, i.e. relatively high (or low) PDI scores.

For more information, see page 8 of the Values Survey Module at http://www.laits.utexas.edu/orkelm/kelmpub/VSM2013_Manual.pdf.

Hofstede Model is a mean of comparison. It doesn't say much for a single National Culture unless it is compared to another or some others.  You can put any value for the C as long as you do it for both the countries you conduct the study to.  Though this would make it hard to plot and compare with existing Hofstede database.  Since it is mend to have scores between 0 and 100, a C of 50 could be astute.

how to calculate the mean score in spss for each question.