Chapter 4: Research Methods in Cultural Psychology
Conducting Cross-Cultural Research
The central challenge: understanding how culture shapes the way people think, using methods that work across diverse cultural contexts.
Example: Patricia Greenfield studied Zinacantecans in Mexico by interviewing with surveys validated in WEIRD populations; the Zinacantecans expected conversational turn-taking, so the survey format (independent, non-conversational questions) caused anger and misunderstandings. This highlights the mismatch between standard survey methodology and cultural conversational norms.
Two guiding themes of cultural psychology research:
Demonstrating cultural similarities (universal psychological tendencies).
Demonstrating cultural differences (culturally shaped tendencies).
For processes with differences, studies aim to understand how people’s cultural experiences produced those differences.
Selecting Cultures to Study
A common approach is to select cultures based on a theoretical variable of interest (e.g., collectivism).
If cultures differ on collectivism and differences emerge in relationship views, this supports a role for collectivism; non-differences suggest weaker or no involvement.
To study universality, compare cultures that vary greatly on multiple dimensions (language, geography, philosophy, education, social practices).
If a similarity emerges despite great cultural difference, this supports universality (example: mentalizing ability across cultures).
Case example: mentalizing studied in Western children vs Baka (Cameroon) children; despite vast cultural differences, both groups performed similarly on mentalizing tasks, supporting an accessibility universal.
Cultural groups: Western children vs Baka children (nonliterate hunter-gatherers in SE Cameroon).
Concept: “accessibility universal” = a process that develops similarly across diverse cultures.
Making Meaningful Comparisons
After selecting cultures, design a study that yields fair, interpretable contrasts.
Learning about cultures before study:
Western psych researchers often rely on introspection, observations, and intuition; cultural psychologists studying another culture face challenges in generalizing their own experiences.
Richard Shweder’s anecdote: Scandinavian researchers studying Indian rural “family meals” assumed a universal pattern that didn’t exist culturally, leading to misinterpretation. The moral: know something about the culture before drawing conclusions.
Ways to learn about another culture:
Read ethnographies and existing texts (rich descriptions; limited by author’s focus).
Collaborate with cultural insiders or researchers from the culture (International Association of Cross-Cultural Psychology as a resource).
Immersion and firsthand experience (highly informative but time-consuming and costly).
Best approach: combine multiple approaches (texts, collaboration, immersion) to gain a culturally informed perspective.
Contrasting Highly Different and Similar Cultures
Survey methodology requires participants to share implicit understandings of the process; cross-cultural surveys face a challenge known as methodological equivalence: ensuring questions are understood the same way across cultures.
When cultures differ greatly in familiarity with research settings (e.g., subsistence cultures vs WEIRD university students), adapting procedures is often necessary; there is a trade-off between experimental control and comparable meaning.
Why most cross-cultural research uses WEIRD samples (university students in industrialized societies):
Generalizability concerns (do findings extend beyond student populations?).
Statistical power concerns (ability to detect cross-cultural effects varies with the degree of contrast between cultures).
If there are pronounced differences, stronger effects are more likely to emerge with more divergent cultures; differences between WEIRD populations are conservative tests of cultural influence.
Surveys and Questionnaires
Cross-cultural surveys face multiple challenges: translation, response biases, reference group effects, deprivation effects.
Translation of Questionnaire Items
Language differences: options include keeping materials in original language (e.g., English) and testing bilingual participants, or translating materials.
Pros of bilingual-only approach: avoids translation issues; cons: participants may be less proficient or unrepresentative; language can shape thought.
Language and thought: bilingual participants may respond differently when tested in native vs second language (language of testing can affect thinking styles).
Translation difficulties: many concepts/terms have no direct equivalents (e.g., amae, schadenfreude, self-esteem in Chinese).
Accurate translation: ideally, a bilingual investigator ensures translations capture subtle meanings and can compare translated materials to originals and back-translations.
Back-translation: English materials translate into another language, then back into English by a second translator; compare originals and back-translations to resolve ambiguities; drawback: back-translation can produce awkward phrasing that is unnatural in the target language.
Translation challenges can lead to misinterpretation; use consensus among bilingual investigators to ensure meaning is preserved.
Response Biases
Even with good translations, responses can be biased by culture-specific tendencies.
Socially desirable responding: respondents try to present themselves in a favorable light; may differ by culture (e.g., modesty vs confidence valued in different groups).
Example: leadership self-assessment could reflect cultural norms about modesty or confidence rather than true leadership ability.
Addressed by designing studies that measure the construct indirectly or with behavioral measures.
Moderacy and extremity biases (response styles): some cultures tend to choose midpoints or extreme ends of scales more than others.
African Americans and Hispanic Americans more prone to extreme responses than European Americans; East Asians tend to be more moderate, and influence of native language can increase moderacy bias.
Mitigation strategies: use yes/no formats to remove middle options; or standardize data (z-scores) to remove response style effects.
Acquiescence bias: tendency to agree with items; controlled by including reverse-scored items and/or standardizing data.
Reversing half the items prevents overall bias from inflating scores.
Standardization (z-scoring) can also mitigate acquiescence but may obscure between-culture mean differences.
Reference group effects: responses can depend on the reference group used for comparison (e.g., height self-ratings vary by culture due to the reference standard).
Consequence: comparing across cultures with a single standard is problematic.
Addressed by using concrete, context-specific items and avoiding subjective, culturally relative standards when possible.
Deprivation effects: cross-cultural values measures may reflect conditions of safety and resources rather than true importance of values (e.g., valuing safety more in high-risk contexts can distort comparisons).
Use converging evidence from other sources besides self-report to validate value measures.
Reducing Biases and Analyzing Data
Use concrete measures instead of abstract, subjective items when possible (e.g., describe actions rather than generic traits).
Use forced-choice response formats to reduce interpretive variance.
Employ behavioral or physiological measures (where feasible) to validate self-report findings and reduce reference-group biases.
Standardization caveat: preserves patterns but may erase cross-cultural differences in average levels on single measures.
The Experimental Method in Cross-Cultural Research
Experiments manipulate an independent variable (IV) and measure its effect on a dependent variable (DV) under controlled conditions; in cross-cultural work, culture is a non-manipulated IV (a quasi-experiment).
Two main manipulations:
Between-groups manipulation: different groups experience different IV levels; requires random assignment to conditions to ensure equivalence.
Within-groups manipulation: the same participants experience multiple IV levels; different order of conditions can be counterbalanced.
Experimental designs help isolate effects of the IV while reducing confounds, and they allow comparisons within cultures (where response biases are shared) to be more valid than cross-cultural mean comparisons.
Example usage: test how different status levels influence evaluations within Jamaicans and Indians, then compare the pattern of effects across cultures. This shifts from comparing means across cultures to comparing patterns of means, which mitigates response biases.
Neuroscience Methods
Cultural neuroscience uses brain-imaging and related methods to examine how neural activity relates to cultural traits.
Examples:
fMRI: study brain regions activated during tasks and compare across cultures; e.g., Telzer et al. (2010) found Latin Americans showed greater reward-system activity when earning money for family vs self compared to European Americans, indicating stronger interdependence.
EEG: measures time course of brain activity in response to culture-relevant stimuli; provides precise temporal information but less spatial localization.
Other neuroscience approaches include molecular genetics to examine how gene frequencies relate to thinking styles across populations.
Neuroscience methods are less susceptible to response biases that bias self-reports and can reveal neural mechanisms underlying cultural differences.
Replication
Replication strengthens confidence in findings; failure to replicate can be due to unreliability in the original, issues in replication, or cross-cultural factors altering the phenomenon.
In cultural psychology, failed replication can also indicate that cultural processes produce different results in different contexts or over time (cultural change).
The field emphasizes replication more now (Open Science; etc.).
Specific Methods for Studying Culture
These methods go beyond standard psychology to address culture-specific questions and constraints.
Situation Sampling
Addresses the limitation that researchers cannot randomize cultures; instead, sample situations that are routinely experienced in different cultures and ask participants to imagine how they would respond.
Two-step procedure:
1) Participants from at least two cultures describe situations in which a specific outcome occurred (e.g., self-esteem increased/decreased).
2) Participants from each culture imagine how they would feel in the other culture’s listed situations, to compare responses across cultures.Example (Kitayama et al., 1997): Americans and Japanese described self-esteem-increasing/decreasing situations; step-2 participants imagined the other culture’s situations; results showed culturally particular self-esteem dynamics.
Analyses enable two types of insights:
Within-culture differences in responses to the same set of situations indicate habitual patterns.
Differences in responses to culturally generated situations indicate different cultural experiences shaping thinking.
Cultural Priming
Priming activates cultural ideas to temporarily shift thinking and behavior toward those associated with a culture.
Finding: thought patterns can be nudged toward independence (American) or interdependence (Chinese) depending on priming; e.g., independence primes shift Chinese self-descriptions toward American styles and interdependence primes shift Americans toward Chinese styles.
Practical note: priming demonstrates that many cultural differences are degree-based rather than kind-based, with nonuniversals (non-generalizable to all contexts) existing in some processes.
Applications: priming methods have been used to explore a broad range of cultural cognition and behavior; future discussions in related chapters.
Contrasting Cultures by Categories
Cultural categories used to compare thinking include:
Tightness–looseness (Gelfand et al., 2011): strength of norms and tolerance for deviance.
Hofstede’s dimensions (1980, 2001): individualism-collectivism, power distance, uncertainty avoidance, masculinity, long-term vs short-term orientation.
Triandis’ refinement: vertical vs horizontal dimensions of individualism and collectivism.
Schwartz’s value dimensions (1994; Schwartz & Boehnke, 2004): universalism, benevolence, conformity, tradition, security, power, achievement, hedonic, stimulation, self-direction, etc.
Culture categorization can reveal predictive associations with psychological traits across contexts.
Culture-level measures use more objective datasets:
Human Relations Area Files (Ember, Ember) – ethnographic database (~700 categories).
Standard Cross-Cultural Sample (Murdock & White) – cross-cultural comparisons.
Database of Religious History – history-based cultural comparisons (expanding beyond religion).
Caveat: databases help with comparisons of smaller cultures but are less useful for contrasting modern industrialized societies.
Culture-Level Measures and Messages
Analyzing cultural messages people routinely encounter can illuminate how cultures shape thinking.
Example workflow:
Identify identifiable, quantifiable cultural messages (e.g., magazine ads, laws, fairy tales, sports coverage, web pages, personal ads).
Develop a hypothesis about the prevalence of certain messages across groups (e.g., working-class vs upper-middle-class Americans; resilience vs uniqueness in music lyrics).
Transform raw data into codified categories (coding). Use blinded coders and multiple coders to ensure reliability.
Analyze the prevalence of themes (e.g., Snibbe & Markus, 2005): country music more frequently conveys resilience; rock music emphasizes uniqueness.
Figure 4.6 illustrates the proportion of songs containing resilience vs uniqueness themes in country vs rock music.
The Challenge of Unpackaging
Cultural differences are often packaged wholes; unpackaging seeks to identify the underlying variables that explain observed differences.
Example: Embarrassability and interdependence differ across cultures; researchers first establish that Japanese are more interdependent and more embarrassable, then examine whether interdependence explains embarrassability by correlating the two within each culture.
If a correlation exists within cultures, it supports the idea that one underlying variable (interdependence) partly explains the observed difference in embarrassability.
Unpackaging helps identify which cultural experiences are most relevant and how they drive cognitive or affective processes.
Using Multiple Methods
No single study is perfect; converging evidence from multiple methods strengthens confidence in findings.
Occam’s razor principle: prefer simpler explanations; when four studies using different methods converge, the single explanation is more convincing than multiple competing accounts.
A well-supported research program combines multiple methods to test a theory from several angles.
Case Study: The Culture of Honor in the Southern United States
Researchers: Nisbett and Cohen explored why the U.S. South shows higher violence than the North.
Historical observations: Victorian-era accounts of violent behavior, lynchings, duels, and cruel practices; the South more tolerant of capital punishment and gun ownership; contrast with Northern norms.
Contemporary indicators: Southern high schoolers more likely to report carrying a weapon; more school shootings in the South; higher domestic violence rates in the South.
The culture-of-honor hypothesis: the Southern settlers were primarily herders (Scotch-Irish borderlands) vs Northern farmers. Herders’ livelihoods were portable and vulnerable to theft; reputational concerns led to a culture where honor needed defending through aggression.
Evidence from multiple methods (case-based, diverse methods) strengthens the theory:
Archival data: homicide patterns show more argument-related killings, especially in rural South; within rural areas, homicide rates are higher in livestock regions than crop regions; argues against poverty or temperature as primary drivers.
Archival data extension: Grosjean (2014) found Scottish-Irish settler regions in the South show higher modern murder rates; stronger formal institutions elsewhere reduce these disputes, supporting the culture-of-honor mechanism where weak formal institutions necessitate self-help disputes.
Survey data: Cohen & Nisbett (1994) found Southerners more likely to justify lethal responses to defend family/honor in certain scenarios; not more likely to justify violence in general.
Testosterone response: Cohen et al. (1996) insult experiment with male students from North vs South; Southern participants showed a marked testosterone increase after insult, while Northern participants did not; supports anger and readiness for aggression in honor contexts.
Behavioral measures: a follow-up experiment used a staged hallway confrontation (a confederate blocking the path) to measure how long participants maintained distance before yielding; Southerners displayed more aggression after insult (shorter yielding distance compared to control) than Northerners; Northern responses remained relatively unchanged.
Field experiments: mail-in “honor letter” vs “control letter” to employers; Southern employers showed more warmth to an applicant who described an honor-related offense, while Northern employers did not show such regional bias; supports situationally triggered honor-retaliation tendencies outside the lab.
Additional methods: situation sampling, cultural priming, and agent-based modeling (computer simulations) show the culture-of-honor emerges as a stable strategy under rugged environments and weak policing; similar patterns found in Turkey, Brazil, parts of Africa, Spain, gangs, the Arab world, and inner cities, suggesting broader applicability to environments with strong honor norms and limited formal dispute resolution.
Overall conclusion: A robust cross-method program supports the culture-of-honor explanation for Southern violence, illustrating the power of using multiple methods to test a complex cultural hypothesis.
Chapter Review (Key Points)
Cross-cultural psychological research requires careful culture selection based on theoretical variables and pre-learning about each culture.
Methodological equivalence ensures participants interpret questions and situations similarly across cultures.
Accurate translation of materials is essential when comparing speakers of different languages.
Cross-cultural survey research must contend with translation issues, response biases (social desirability, moderacy/extremity, acquiescence).
Reference group effects arise when self-evaluations depend on culture-specific comparison standards.
The experimental method is valuable for cross-cultural studies because it can control for response biases and reference-group effects by enabling within-culture comparisons and cross-cultural pattern analyses.
Cultural neuroscience (e.g., fMRI, EEG) provides a growing set of tools to examine neural correlates of cross-cultural thinking, with advantages in reducing self-report biases.
Replication strengthens reliability; cultural findings may fail to replicate due to culture-specific processes or changes over time.
Situation sampling provides a practical method for approximating experimental manipulation of culture by analyzing culturally regular situations across groups.
Cultural priming demonstrates that activating cultural ideas can shift thinking toward certain cultural mindsets.
Categorizing cultures along dimensions (tightness–looseness, individualism–collectivism, long-term vs short-term orientation) helps predict and understand cultural differences across contexts.
Culture-level measures (HRAF, SCSS, religious history databases) enable comparisons across smaller societies and across historical periods.
Studying cultural messages (e.g., lyrics, ads, laws) through coding can reveal the kinds of messages people routinely encounter.
Unpackaging identifies underlying variables that explain cultural differences, often via correlational or within-culture analyses.
Using multiple methods yields convergent evidence and strengthens explanatory power; Occam’s razor favors simpler, unified explanations.
The Culture of Honor case study exemplifies how a multi-method program can support a culturally grounded theory of violent behavior, integrating archival data, surveys, physiology, behavioral experiments, and field studies.
Overall, culture shapes thinking in nuanced, often degree-based ways; practical methods like priming, situation sampling, and neuroscience provide powerful complements to traditional surveys for cross-cultural research.