Table 1: Confirmatory Factor Model based on Polychoric Correlation Matrix
Observed Indicators
Factor 1
Factor 2
Reliability
Satisfaction with Democracy
.681
–
.464
(.019)
–
Rely on Parliament
.603
–
.364
(.018)
–
Rely on Government
.615
–
.365
(.018)
–
Membership in EU Good
–
.785
.617
(.016)
Country Benefit from EU
–
.996
.992
(.020)
Factor Correlation = .386
χ
2
= .41, df = 2
N = 16,701
4
Modelling Expressions of Support for the European Union
In order to test the effects of information processing and evaluations of national level institutions
on specific support for the EU I draw on several questions from Eurobarometer 52.0. First, to
measure information processing I use the ability of respondents to answer factual questions about
international and EU politics.
Q44. “Now let us look at some public figures. Could you give me the name of the individual who
currently holds the following jobs?” 1. President of the European Commission; 2. European Com-
missioner appointed by [national government]; 3. [Nationality] Minister of Finance; 4. [Nationality]
Minister of Foreign Affairs.
The variable “Know Facts” is the sum of correct answers to these questions. Because these
questions ask respondents to demonstrate their knowledge of EU and national politics, it is a better
gauge as to whether or not respondents are capable of processing abstract political information than
levels of education. I also include a control for education.
D8. “How old were you when you finished studying?” 1. Up to 15 years; 2. 16-19 years; 3. 20
+ years; 4. Still studying.
Because levels of political discussion have been used as a proxy measure for cognitive abilities I
include a control for this variable as well.
Q5. “When you get together with friends would you say you discuss politics frequently, occa-
sionally, or never?” Answers recoded 1 = never, 2 = occasionally, 3= frequently.
13
Convention