Post on 01-Apr-2015
Cross Cultural Research
Dr. K. A. KorbUniversity of Jos
15 May 2009
OUTLINE
• Issues in Cross-Cultural Research– Validity– Participant– Instrument
• Types of cross-cultural Research
Dr K. A. KorbUniversity of Jos
Validity
• Validity: The degree of accuracy with which a conclusion is drawn – Researchers have a burden to demonstrate that
their research finding have validity– The validity of a study’s finding is directly related
to the validity of the research methods• Interpretive validity: Valid communication
between the research and target group
Dr K. A. KorbUniversity of Jos
Validity
• Ecological validity: Extent that findings generalize beyond the research procedure to the natural context
• Theoretical (construct) validity: Establishing valid operational measures for the concept being studied– This related to:
• Independent variables• Dependent variables
Dr K. A. KorbUniversity of Jos
Research Example
IV: RegularReading
DV:Reading
Achievement
Read bookevery day with
an adult for10 min.
Scores onreading sectionof classroom
exams
Theory:
Research Study:
IV: RegularReading
DV:Reading
AchievementTheory:
Dr K. A. KorbUniversity of Jos
Validity
• Equivalence: For a cross-cultural comparison to be meaning, two condition must be met:– Equivalence in the conceptual meaning of the theoretical constructs
between cultures– Equivalence in the relevance of the empirical method between
cultures
• If any aspect of a research study is not equivalent across cultures then the comparison not valid
Dr K. A. KorbUniversity of Jos
Validity
• Types of Equivalence – Theoretical:Equivalent in meaning of theoretical framework tested
and specific hypotheses tested – Linguistic: Words used in the instruments are similar across languages– Measurements: Instruments are equally valid and reliable across
cultures– Sampling: Samples are representatives of their culture and equivalent
on noncultural demographic characteristics– Procedure: Procedures used to collect data across cultures are similar
Dr K. A. KorbUniversity of Jos
Participants
• Cultures to be studied– Differ on independent variables of interest– Check to confirm that the cultures do differ on the
independent variable
• Subgroups within the cultures– Should be as similar as possible on non-cultural relevant
demographic characteristics
• Individuals within the subgoups
Dr K. A. KorbUniversity of Jos
Instruments
• Perspective 1: Use similar instruments for each culture– Assumption: The instrument is not biased against
a particular culture• Conclusion: Differences between groups reflect
construct differences in each culture – Advantage: Comparison is straight forward– Disadvantage: The instrument might not be
culturally meaningful for all cultural groups
Dr K. A. KorbUniversity of Jos
Instruments• Perspective 2: Use different instruments for each
culture– Assumption: The same instrument cannot be equivalent
across culture because psychological constructs are context-bound• Conclusion: Difficult interpretation on comparisons across cultures
– Advantages: Measures behaviors that are more appropriate for each culture
– Disadvantage• Uncertain whether the instruments are measuring conceptually
similar constructs• Comparison between cultures is difficult
Dr K. A. KorbUniversity of Jos
Instruments
• Linguistics Equivalence: Can be established through back-translation of an instrument– Step 1, Forward Translation: Translate from English into
Language– Step 2, Back Translation: An independent person
translates the translation back into English.
• The instrument is considered equivalent in both languages if the back-translation is equivalent to the original
Dr K. A. KorbUniversity of Jos
Instruments
• Cultural Bias: The instruments does not measure equivalent concepts in both cultures– Theoretical/construct bias: The definition of the concept differs
between cultures– Item bias: Poor item translation – Procedural bias: Differences in testing procedures across cultures
• Testing conditions• Family with testing procedures• Response styles
Dr K. A. KorbUniversity of Jos
Instruments
• Response Style: Systematic tendency to respond in a certain way to items on a questionnaire– Acquiescence bias: Agree with statements as presented – Central tendency bias: Avoid extreme categories – Social Desirability: Tendency to portray oneself in favorable light– Reference Group Effect
• Participant self-report by comparing themselves to others• Individuals with different cultural will have different standards of
comparison
Dr K. A. KorbUniversity of Jos
Reference Group Effect
• Standard item: I have respect for the authority figures with whom I interact.
• Cultural comparison item: Compared to most Japanese I know, I think I have respect for the authority figures with whom I interact.
Dr K. A. KorbUniversity of Jos
Reference Group Effect
Dr K. A. KorbUniversity of Jos
Control of External Variables
• To reduce alternative explanations of differences between cultural groups– Cultural populations should be selected a priori
(before the experiment) based on ethnographic descriptions
– Measure the dependent variable by two or more measures
– Eliminate effect of irrelevant variables though statistical analysis (ANCOVA)
Dr K. A. KorbUniversity of Jos
Research Designs
• Causal- Comparative: Compare groups of people on an independent variable to determine the effects of an independent variable that cannot be manipulated
• Correlational: Compare one group of people on multiple variables to determine the relationship between variables.
• Experimental: Compare treatment and control groups on a dependent variables. The experimenter must implement a treatment.
Dr K. A. KorbUniversity of Jos
Cross-Cultural Research• Cross- Cultural Comparisons: Studies comparing different
cultures on the same psychological variable– Determine how individuals within cultures differ on psychological
variables of interest• Cultural Studies: Studies based on a theoretical framework
that predicts and explains differences between cultures– Correlation studies: Studies that determine whether the relationships
between psychological variables are similar across cultures• Linkage studies: Studies that establish links between contents
of culture and psychological variables of interest – Unpacking studies– Experiments
Dr K. A. KorbUniversity of Jos
Unpaking Studies
• Instead of culture as independent variable, a Context Variable is hypothesized to influence the dependent variable – Context variable: Specific variable that explains cultural
differences that then influences the dependent variable• Both the context variable and the dependent
variable are measured– Degree to which the context variable influence the
dependent variable is statically tested
Dr K. A. KorbUniversity of Jos
Cross-Cultural Experiments
• Priming Experiment: Experimentally manipulate mindsets related to culture to determine whether behavior differs as function of a primed mindsets
• Example (Trafimow,Triandis,& Goto, 1991)– Treatment 1: Private Mindset
• Please think of what makes you different from your family and friends.– Treatment 2: Collective Mindset
• Please think of what you have in common with your family and friends. What do they expect from you to do?
– Participants were then asked to write a paragraph that described themselves
Dr K. A. KorbUniversity of Jos
Frequency of Individually-Oriented and Group-
Oriented Responses for American and Chinese (Trafimow et al., 1991)
Dr K. A. KorbUniversity of Jos
Conclusion
• Only compare populations that have theoretical reasons for differing on a dependent variable
• Replace culture with specific independent variables that might influence the dependent variable– What aspects of the cultures cause differences in
the dependent variable?
Dr K. A. KorbUniversity of Jos
Revision
• What validity consideration need to be taken into account in cross-cultural research studies?
• What are some key issues to consider when developing instruments in cross-cultural research studies?
• Describe research designs that are useful in cross- cultural research studies
Dr K. A. KorbUniversity of Jos