br 
multicultural research, not to confirm the hypothesis, but to understand it in context with other 
hypotheses. 
Multimethod research is particularly useful in coping with the various forms of artifacts 
that threaten social! psychological research generally, and cross-cultural research in particular. 
For example, it has long been argued that experimenter expectancies and demand ] 
characteristics, positive and negative subject roles, and sampling bias threaten the validity of 
social psychological research (e.g. Adair, 1973; Orne, 1962; 1969; Rosenthal, 1966; Silverman, 
1977; Weber & Cook, 1972). In cross-cultural research, the difficulties include lacking control of 
cultural variables, establishing conceptual and metric equivalence, disentangling emics and 
etics, and choosing representative versus equivalence sampling (e.g. Berry, 1969; 1980a; Bijnen, 
Van Der Net & Poortinga, 1986; Lonner & Berry, 1986; Malpass & Poortinga, 1986). 
Given that all humans are reactive, semi-autonomous agents and that culture ubiquitously 
influences all aspects of human thought and behavior, including that of the researcher, it is 
indeed difficult to conceive and execute a flawless cross-cultural, social psychological study. 
However, the multimethod research advocated by Rohner (1977), incorporating “constructive” 
replication (Lykken, 1968) and muitiple sampling at different levels of human social behavior, 
offers some surity. Acknowledging that every research method has strengths and weaknesses 
and potential for systematic error and bias, surity is sought by triangulating multiple research 
methods and anticipating that converging results will counteract doubts arising from the 
potential design weaknesses of any one study (Rosnow, 1981; Runkel & McGrath, 1972). 
This thesis research design was rather directly modeled on Rohner’s recommended 
triangulation scheme, but with accommodation to the goals of generative research. Thus, three 
types of studies were undertaken to test the hypothesis that ownership reflects interpersonal 
dominance and to explore the meaning of that hypothesis in context with competing 
hypotheses. The triangulated methods were 1) holocultural research, 2) psychological research, 
and 3) cross-cultural field research. Hoiocuitural research samples panculturally, using 
archived data that were not collected by, or for the purposes of, the researcher. Since cultures 
are used as-cases, holocultural research allows interpretations in terms of cultural ecology 
(Berry, 1979; Hofstede, 1980). Psychological survey research examines individual differences 
across relatively large numbers of people within a single culture, thus holding cultural variables 
constant. Cross-cultural field research examines differences within and across relatively small