Methods and Bias Flashcards
Types of studies
- Exploratory studies
- Hypothesis-testing
Exploratory studies
Examines the existence of cross-cultureak similarities or differences
- strength: broad scope for identifying similarities and differences
- weakness: limited capability to solve the causes of differences
Hypothesis-testing studies
Examines why cultural differences exist
- inferences promote cross-cultural biases and inequivalence
What are we looking for with studies?
- structure-oriented studies
Structure-oriented studies
Comparison of a culture’s constructs structures, or relationships amoing constructs with those of another culture
At what level are we doing research?
- individual level studies
- ecological (cultural) studies
- multilevel studies
Individual level studies
Individual participants provide data and are the unit of analysis
Ecological (cultural) studies
Countries or cultures are the units of analysis
- Hofstede
Multilevel studies
Involve data collection at multiple levels of analysis
Distinguising levels is improtant
Because of isomorphism (the realtionship of the variable is the same across levels)
- but this is not a given: relations between variables may not be the same at different levels
The Simpsons paradox
The relation between two factors on an individual level can be negative while the relation an the national level is positive. So the effect of the relation differs across different levels
- obese and health individually is bad, but national means more wealth and healthcare
Challanges when designing studies
Are the differences actually cultural –> validity of theoretical models
- Linkage studies
Linkage studies
- Unpackaging studies
- experiments
Unpackaging studies
Unpackaging studies seperate the contents of the global, unspecific concept of culture into specific, measurable psychological constructs and examine their contribution to cultural differences
- lots of contextual information is important
Poortinga on culture
Culture as a placeholder for what we do not understand yet
–> Culture is like peeling an onion: peeling away layer by layer
Experiments (linkage studies)
Priming studies and behavioral studies
Priming studies
Involve experimentally manipulating the mindsets of participants and measuring resulting chamges in behavior
Behavioral studies
Involve manipulations of environments and observation of changes in behavior
The ecological fallacy
Inferences on individuals based on aggregated country-level data
–> make a statement about an individual based on the mean of the country
(Differences within groups can be bigger than differences between groups)
Bias
Differences that do not have the sam meaning within and across cultures
- if bias exists in cross-cultural comparative study, the comparison loses its meaning
Equivalence
State or condition of similarity in conceptual meaning and empirical method between cultures
- allows comparisons to be meaningful
–> what is the level of comparability
Measurement bias
Are the specific measures, tests, or instruments used to collect data in different cultures equallu valid and reliable across those cultures?
Construct bias
Are psychological constructs defined in the same way across cultures>
Linguistic bias
re the research protocols semantically equivalent across the languages used in the study?
Response bias
Do people of the different cultures have different tendencies to respond to questions?
Model bias
Do the theoretical framework and hypotheses being tested mean the same thing in the cultures being tested?
- happiness; is it same everywhere?
Sampling bias
Are the samples in the cultures tested appropriate respresentative of their culture and equivalent to each other?
Procedural bias
Do the procedures by which data are collected mean the same in all cultures tested?
Interpretational bias
Are statistically significant findings practicually meaningful?
- are the interpretations made about the findings and conclusions drawn biased in some way?
- are the interpretations about cultural sources of differences justified by data?
Filial piety
- Example for construct bias
Respect for parents is universal
But: it can be more or less extensive, for instance when deciding what a family does when their elderly parents need support
Item/linguistic bias
An item is biased when it has a different psychological meaning across cultures
- item is not applicable
- has cultural connotations
- cannot be translated
Different method biases
- Measurement/instrument bias = test characteristics
- Linguistic/item bias = are items semantically equivalent?
- response bias = response styles (extremities or acquiescence)
- sampling bias = confounding differences
- procedural/adminsitration bias = procedural (interviewer effects)
Test adaption for translations
Forward and backward translation: translate from on to another language and someone else translate it back, to see if the interpretation is the same
- committee approach
- cognitive interview
Interpretational bias
- Poortinga’s ways for dealing with nonequivalence
- ignore the nonequivalence
- interpreting the nonequivalence
- preclusing comparison
- reducing the nonequivalence in the data
Method validation and indigenous studies
Cross-cultural validation studies
Indigenous cultural studies
Cross-cultural validation studies
studies that examin whether a scale, test, or measure originally developed in a culture is valid in another culture
Indigenous cultural studies
Rich description of complex theoretical models within a single culture
- insights generated from these studies compared across studies and cultures
