Page 124 - 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

bias (see Hui and Triandis 1989; Cheung and Rensvold 2000; Little 2000
for publications on “response sets”; Bortz and D¨ oring 2002, 215).
Adminstration equivalence and administration bias: Bias on the ad-
ministration level can occur due to culture-specific attitudes of the in-
terviewers that might produce culture-specific answers. Another source
of administration bias could be found in sociodemographic differences
betweenthevariousnationalinterviewerteams(vandeVijverandTanzer
1997, 264).
It is noteworthy that the functional equivalence approach to interna-
tional research overrides the inaccurate belief that identical instruments
automatically measure in an identical or equivalent way. For example,
Greenfield (1996, 330) emphasizes the great potential of using video
recording in process-oriented research on cultural differences without
expounding on the problems caused by the culture-specific reaction of
being recorded. Such a reaction would most probably vary between cul-
tures in the field of participant-observation studies, too. A Caucasian
observer would not be the cause of any bias in Europe, but might
cause a sensation in rural parts of Africa or Asia (Niedermayer 1997,
93–7).
Method bias is especially treacherous, because it will appear as
“cultural differences” in the results when analyses of variance are under-
taken. These differences are bound to run directly into misinterpretation
due to their source of inadequate measurement. Moreover, significant
cultural differences could be lost in a mixture of method bias and “real”
results, so that the entire interpretation would be useless (van de Vijver
and Leung 1997, 15–17).

TESTING FOR AND ESTABLISHING EQUIVALENCE:
A GUIDELINE

As shown previously, equivalence can be seen as the major problem of
comparative research. For data to be adequate and interpretable, the re-
searcher requires the highest possible level of equivalence. A hierarchical
step-by-step procedure can help to test for and establish equivalence on
every level of the research process. The following guideline will describe
the most important steps in the establishment of equivalence. 4

4
See van de Vijver and Leung (1997, 42–51) and van Deth (1998, 9) for other guidelines.

104