Page 122 - Comparing Political Communication Theories, Cases, and Challenge

P1: JZZ/KAA P2: KaF
0521828317c05.xml CY425/Esser 0521828317 May 22, 2004 11:55

Werner Wirth and Steffen Kolb

indispensable, especially if the research team is nationally based and is
not cooperating with international partners.
It is extremely difficult to prove construct equivalence beyond any
doubt. Construct validation is a difficult task in one country alone, so
that the international perspective only complicates the situation further.
Besides, an identical state of internal and external structures might even
be the result of cultural bias and/or occur at random. The probability
of a random state of “sameness” decreases with the number of countries
included in the study. Moreover, this probability can be calculated sta-
tistically and published in the findings. Most procedures of construct
validation require multidimensional or item-battery measurement, be-
cause the internal structure cannot be tested when measured in just one
variable (van de Vijver and Leung 1996, 273; 1997, 17; van Deth 1998). In
the face of these problems, one might ask whether the effort undertaken
to achieve construct validation is worthwhile. Yet, only the procedures
we have presented will lead to a well-founded decision, irrespective of the
preferred approach to be undertaken for the enquiry, be it etic, emic, or
etic-emic (van de Vijver and Leung 1997, 12–15).

Item Equivalence and Item Bias
Even withagiven construct equivalence, bias can still occur on the
item level. The verbalization of items in surveys (and of definitions and
categories in content analyses) can cause bias due to culture-specific con-
notations. Item bias is mostly evoked by bad, in the sense of nonequiva-
lent, translation or by culture-specific questions and categories (van de
Vijver and Leung 1997, 17). Psychological inventories and item batteries
in particular can be tested for item equivalence, using several procedures
derived from the “item response theory” (e.g., Lienert and Raatz 1994;
van de Vijver and Leung 1997, 62–88). Compared to the complex proce-
dures discussed in the case of construct equivalence, the testing for item
bias is rather simple (once construct equivalence has been established):
Persons from different cultures, who take the same positions or ranks
on an imaginary construct scale, must show the same attitude toward
every item that measures the construct. Statistically, the correlation of
the single items with the total (sum) score have to be identical in every
culture, as the test theory generally uses the total score to estimate the
position of any individual on the construct scale. Hui and Triandis (1985,
135) add scalar equivalence to the list. This tests whether the construct
is measured on the same scale. In general, different wordings of ques-
tions or categories have to be well-founded and published in the research

102