How to perform mixed model analysis for my data in SPSS?

Question

In my thesis, I am trying to find out what factors influence the behavior of corporate (corporate social responsibility, GSE_RAW ) companies. Two groups of possible factors / variables were identified: country-specific and country-specific.

First, company-specific variables are (among others)

• MKT_AVG_LN : company market value
• SIGN : number of CSRs signed by the company
• INCID : the number of CSR incidents in which the company participated in

Secondly, each of the 4000 companies in the data set has a head office in one of 35 countries. For each country, I collected some data for specific countries, among other things:

• LAW_FAM : the legal family with which the legal system of countries is associated (French, English, Scandinavian or German).
• LAW_SR : relative protection granted to company shareholders (for example, in case of default of the company).
• LAW_LE : relative efficiency of the legal system of countries (higher value means more effective, for example, less damaged).
• COM_CLA : measure the intensity of internal market competition.
• GCI_505 : measuring the quality of primary education
• GCI_701 : measuring the quality of secondary education
• HOF_PDI : power distance (higher value means more hierarchical society).
• HOF_LTO : country orientation (higher means longer term orientation)
• DEP_AVG : GDP of countries per capita
• CON_AVG : average inflation of countries compared to the period 2008-2010

To analyze this data, I “raised” the data at the country level at the company level. For example, if in Belgium the value of COM_CLA is 23, then all Belgian companies in the dataset have a value of COM_CLA of 23. The variable LAW_FAM is divided into 4 dummy variables ( LAW_FRA , LAW_SCA , LAW_ENG , LAW_GER ), giving each company 1 for one of these dummies .

All this leads to a data set as follows:

 COMPANY MKT_AVG_LN .. INCID .. LAW_FRA LAW_SCA .. LAW_SR LAW_LE COM_CLA .. etc ------------------------------------------------------------------------------ 1 1.54 55 0 1 34 65 53 2 1.44 16 0 1 34 65 53 3 0.11 2 0 1 34 65 53 4 0.38 12 1 0 18 40 27 5 1.98 114 1 0 18 40 27 . . . . . . . . . . . . . . . . 4,000 0.87 9 0 1 5 14 18 

Here are companies 1 to 3 from one country A and 4 and 5 from country B.

At first I tried to analyze the use of OLS, but the model seemed very “unstable”, as shown below. The first model has an r-square .516:

Adding only two variables changes many levels of beta and significance, as well as the r-square (.591). Of course, the r-square increases when variables are added, but this is a pretty big increase from .516:

In the end, it was suggested in another post that I should not use OLS here, except for mixed models, due to categorical account level data. However, I am confused about how to accomplish this in SPSS. The examples I found on the Internet are not comparable with mine, so I don’t know what to fill out, among others, in the following mixed models dialog:

Can anyone use SPSS, please help me explain how to perform this analysis so that I can come up with a regression model (CSR = b1 * MKT_AVG_LN + b2 * SIGN + ... + b13 * CON_AVG) so that I can conclude that CSR is determined by company features or country characteristics (or neither, nor both)?

I believe that I need to insert company level variables as covariates and country level variables as factors. It's right? Secondly, I'm not sure what to do with dummy variables LAW_SCA to LAW_ENG .

Any help is much appreciated!