Cultural Psychology: Chapter 4 - Methods for Studying Culture and Psychology

Dependent variable

In an experiment, the variable or measure affected by manipulation of the independent variable.

What kind of evidence and sub-studies (my word) did Nisbett and Cohen use/conduct to test their hypothesis? What did they find?

1. Archival data on police records of homicides: accumulated documents or records of a culture. Showed higher argument-related murders, not just homicide rates in general, in the South than in the North, especially in the rural South compared with the rural North. Also, the homicide rate is higher in the herding regions than in the farming regions of the South, and this can't be explained by slavery, temperature, or income. 2. Survey data: attitude items, scenarios regarding the appropriateness of violence. Southerners were only more likely than northerners to have more positive attitudes toward violence when it related to defending their families or honor. 3. Physiological measures: Non-Hispanic White male students at the University of Michigan who had grown up either in the North or the South. Going up and then down hallway to pick up a questionnaire. Confederate was angry, on way back, called person "*******" after bumping participant. Dependent variable: change in participants' testosterone level throughout the experiment based on saliva samples. Southerners who had been insulted showed a sharp spike in their testosterone level, whereas northerners didn't show a significant difference in their testosterone level between the two conditions. 4. Behavioral measures (perhaps most exciting for many psychologists): same "*******" manipulation but with an added component. Dependent measure: how close the participant got to the second confederate before yielding away. Northerners showed no sig difference in their behaviors between the insult and control conditions. Southerns showed a pronounced difference. When not insulted, they yielded much earlier than northern subjects (southern hospitality). When insulted, they went right up to the face of the second confederate. (aggression carried over to their interactions with a new person) 5. Field experiment: to see if there are regional differences in people's tolerance for insult-related aggression in the real world. Letters: applicant, "honor letter" (manslaughter, killed man who slept with girlfriend and teased him publicly) "control letter" (served time for stealing a car). Dependent variable: the warmth of the letter that the employer sent in return. Southern employers were more sympathetic to the applicant in the "honor letter" condition than the northern employers, but there were no regional differences in warmth in the "control letter" condition.

Between-groups

A type of experimental manipulation in which different groups of participants receive different levels of the independent variable(s). Requires random assignment so that the participants in the different conditions are statistically equivalent at the beginning of the study.

How can one avoid moderacy and extremity biases?

1. Avoid providing participants with a set of responses that has a middle answer (Yes/No), but this may not be sensitive enough to detect differences. 2. Standardize your data. (standardized scores, or z scores). Participants' responses are now expressed in terms of the number of standard deviations by which they depart from the participant's own personal average. Eliminates the problems with moderacy and extremity biases but also assumes that the average level of response is identical across cultures (average z score of 0 for everyone's responses). Ok when looking at pattern of responses but not when comparing the average level of responses across cultures.

What are some techniques for correcting the problems associated with the reference-group effect?

1. Avoid subjective measures that might have different standards in the groups being compared. Change the content of the item (make it more concrete of a scenario, but this means also that the scenario may be more specific, so you would want to create a number of items) or the response format (quantitative descriptions: at least once a day, 10-20% of the time; forced choice between 2 or more alternatives -- Which of these two responses would you be more likely to choose if you were in such a situation?) 2. Employ behavioral measures that don't rely on people's understanding of how they compare against others. (timing people, physiological measures, like autonomic nervous system response)

What are some strategies to getting over the issue of conducting survey research across cultures? This has to do with language differences, and number 1 should be the author's favored approach.

1. Be sure at least one of the primary investigators on a project if fully bilingual in the languages being compared. 2. Use bilingual participants 3. Use a translator 4. Back-translation (might result in a very unnatural or hard to understand translation, even though the literal meaning is preserved -- "piece of cake")

What kinds of analyses can researchers do with the situation sampling methodology?

1. Can explore cultural differences in response styles: If, regardless of the situation, people in one culture indicate that they would consistently respond differently from those in the other culture, this suggests that certain learned cultural experiences have become habitualized by people, and these learned cultural experiences govern people's reactions across all kinds of situations. (there are certain experiences that differ from culture to culture and influence how people respond to different situations) 2. Researchers can explore whether the cultural origin provided in step one are respnded to differently by participants in step two. If the culture appears to matter in how participants respond (participants respond differently depending on which culture the situation comes from), this would suggest that the two cultures provide participants with different kinds of experiences.

What are the methodologies/strategies proposed by the author to improve the ability to assess psychological states of people from other cultures? This should be a list in chronological order starting from creating the construct.

1. Choose samples based on a theoretical variable that you're investigating. Specifically, look for samples that differ on the theoretical variable/dimension. If similar ≈ high degree of universality. 2. Develop some knowledge abut the cultures under study. (researcher's own experiences may not generalize to the people she's studying; read existing texts and ethnographies; find a collaborator from that culture; immerse yourself in that culture 3. Adapt the procedures so that they are understandable in each culture we study, which sometimes results in our using a slightly different procedure in each culture (methodological equivalence) (standard survey methodologies may not be of much use when exploring subsistence societies that don't share these kinds of research experiences) (trade off: some experimental control is lost, but comparable meaning is "intensified")

What two themes guide this book?

1. Demonstrate similarities across cultures in the ways people think or to demonstrate cultural differences, which reflect culturally shaped psychological tendencies. 2. When cultural differences emerge, understand how people's experiences shape the ways they think.

What steps should be taken when studying cultural messages?

1. Focus our investigation on an identifiable and quantifiable subset of the sources of cultural messages (magazine advertisements, laws, fairy tales, personal ads). 2. Derive a specific hypothesis 3. Come up with a way to transform our raw data into quantifiable data that lend themselves to a test of our hypothesis. This involves coding the data (deriving a set of categories consistent with particular cultural messages, like whether the messages emphasize something relevant to resilience, like rebelling against authority): can be subjective. Safeguards: coders are kept blind to the hypotheses so that they're not motivated to get good results from leaking into the subjective decision they have to make; employing multiple coders and coding some or all of the raw data by more than one individual and comparing if the data are reliable and training the coders then about the boundaries of the categories (data can be statistically analyzed when coders have reached a consensus for all the decisions they've made)

What are the steps used in unpackaging?

1. Let a theory guide the researchers' search for potential underlying cultural variables (maybe a concern with interdependence and how one relates with close others leads one to become more easily embarrassed) 2. Demonstrate an aspect of this theory, like the fact that Japanese really do have more interdependent view of self than Americans (versus just assuming this) (Japanese do score higher than Americans on embarrassability and interdependence) 3. Demonstrate that the observed cultural difference relate to the observed differences in the other variable (correlate; interdependence and embarrassability do appear to be significantly related within both Japanese AND American cultures)

Describe the two-step process of the situation-sampling methodology.

1. Participants from at least two cultures are asked to describe situations they experienced in which something specific has happened, like situations in which the self-esteem of Japanese and American participants increased or decreased. 2. Different group of participants are provided with a list of situations generated by the first set, and they are asked to imagine how they would have felt if they had been in those situations (indicate how much they thought their self-esteem would have increased or decreased in those situations).

What are some challenges the author presents to conducting survey research across cultures?

1. Research participants and the researchers often speak different languages: Bilingual participants may not be best: may not be representative of their cultures, might be more Westernized, the language we're thinking in can affect the way we're thinking -- respond differently when tested in second or first language; Many psychological terms don't have equivalents in other languages, like certain emotion words 2. Response biases (Moderacy bias; extremity bias; acquiescence bias; reference-group effects; deprivation effects (I don't know whether these last two are response biases)

What are some solutions to the acquiescence bias?

1. Reverse-scoring (half of our items indicate low self-esteem and the other half indicate high self-esteem) = neutralize the effects of this bias 2. Standardize the data

What problems emerge when psychologists overemphasize students in their samples?

1. Significant problem with generalizability (less diverse range of samples = less confident about generalizing to populations other than the samples studied) 2. Power: students from industrialized societies share many similar experiences, so a failure to find cultural differences in a phenomenon between such populations does not necessarily mean that the phenomenon is not influenced by culture. On the other hand, cultural differences seen with similar samples may demonstrate that the effects of culture would be at least as pronounced with more divergent samples.

Culture of honor

A culture in which people (especially men) strive to protect their reputation through aggression.

Back-translation

A method of translating research materials from one language to another whereby a translator translates materials from Language A to Language B and then a different translator translates the materials back from Language B to Language A. The original and twice-translated version in Language A are then compared so that any discrepancies between them can be resolved.

Situation sampling

A methodology that utilizes the fact that cultures do not affect us in the abstract; they affect us in particular, concrete ways. Our experiences in culturally shaped situations lead us to adopt habitual ways of thinking about ourselves and our worlds. The underlying idea of this methodology is that if we can see how people respond to situations that are regularly experienced by people in another culture, we can get a viewpoint into how cultures shape our ways of thinking. Involves a two-step process.

Reference-group effect

A tendency for people to evaluate themselves by comparing themselves with others from their own culture. (Canadians, Japanese, and those from the Netherlands would all have different answers to the statement, "I am tall.") (African-American soldiers in the northern US were less satisfied than those in the South because they compared themselves primarily to civilian African American who were better off in the North than in the South). In this situation with the African American soldiers, the reference-group effect leads one to make the exact opposite conclusion than a more objective comparison would indicate.

Deprivation effect

A tendency for people to value something more when it is lacking in their culture. Therefore, it can be problematic to make inferences about a culture from the values that people endorse the most.

Acquiescence bias

A tendency to agree with most statements one encounters. If people in some cultures (East Asians, for example) have a predisposition to see the truth in more statements than those in another culture (holistic way of looking at the world = interconnected place that's always changing), this will lead to cultural differences in responses, independent from the content of the items.

Independent variable

In an experiment, the variable or condition that the experimenter manipulates in order to examine its effect on the dependent variable.

Within-groups

A type of experimental manipulation in which each participant receives more than one level of the independent variable(s). Does not involve random assignment because each participant is assigned to all of the conditions, but it's important to provide participants with different orders of the conditions (not everyone hears the slow-talking salesperson first). In this study, the order of the conditions is a between-groups manipulation.

Archival data

Accumulated documents or records of a culture.

_____ tend to give more extreme responses than ____.

African-Americans and Hispanic-Americans than Americans of European descent; European-Americans than East Asians

What are the two kinds of manipulations of independent variables (note: independent variables) that can be performed in psychological research?

Between-groups manipulation; within-groups manipulation

How can one develop some knowledge about the cultures under study?

Combination of: Read existing texts and ethnographies (almost definitely biased by beliefs, values, etc. of author(s); may not give a lot of information about the variable of interest); find a collaborator from the culture; immerse yourself in that culture (time-consuming, costly)

Ethnographies

Contain rich descriptions of a culture, or a particular situation or group of people within a culture, derived from extensive observation and interaction by an anthropologist.

What is the method of choice for psychological research?

Experiment. However, because one's cultural background isn't manipulated, comparisons of cultures are not true experiments but are quasi-experiments. This can provide us with a comparison not limited by the response biases associated with questionnaire research. We are explicitly comparing, say, Indians with Indians, so response biases are not affecting our comparisons. We can compare the magnitude of the impact of status between cultures. It should be noted that the experimental method can be utilized in questionnaire research whereby the questionnaires have different conditions. The experimental method changes the between-culture comparison from comparing the magnitude of means across cultures (problematic) to comparing the pattern of means across cultures (fine).

Field experiment

Experiments conducted outside the laboratory in the real world, with participants who are typically not aware that they are in a study.

Unpackaging

Identifying the underlying variables that give rise to different cultural differences. Cultural difference don't tell us which cultural experiences sustain them. Cultures appear to us as "packaged." Broad network of practices and meanings that are hard (or impossible?) to separate. Does not necessarily mean that you've identified all the variables behind a cultural difference or even the most powerful variables. This can help psychologists learn more about particular psychological phenomena, which helps us understand our own minds (interdependence and level of embarrassibility are related to one another, which may have certain implications about the cultures and our own minds). Allows for the identification of other psychological processes that relate to the cultural differences that have been observed.

Methodological equivalence

In cross-cultural research, the concern with making sure participants from different cultures understand the research questions or situations in equivalent ways.

Culture-level measures

It's important that we have a confident understanding of what the cultures are like in the first place (we need a way to measure cultures). A careful analysis of the messages that people are exposed to on a regular basis can provide us with a nice perspective on the ways that cultures influence their members (magazine advertisements, laws, fairy tales, personal ads).

What did a study using the situation sampling methodology demonstrate about Japanese and Americans?

Japanese participants' self-esteem tended to increase less than that of Americans in self-esteem-increasing situations, and their self-esteem decreased more than that of Americans in self-esteem-decreasing situations. (Habitually more attentive to situations that afford opportunities for self-criticism, whereas Americans are habitually more attentive to situations that boost the positivity of their self-views.) American and Japanese participants in step two reported that their self-esteem would decrease more when responding to Japanese-made self-esteem-decreasing situation than American-made ones. Self-esteem would increase more when responding to American-made self-esteem-increasing situations than Japanese-made ones. This suggest that the kinds of experiences people encounter in the US are especially conducive to boosting self-esteem, whereas the kinds of experiences people regularly encounter in Japan lead people to be especially self-critical.

What are the response biases the author gives?

Moderacy bias; extremity bias; acquiescence bias; reference-group effects; deprivation effects (I don't know whether these last two are response biases)

Argument-related murders

People are compelled to defend their honor

Who studied the question, "Why does the US South seem so much more violent than the North?"

Richard Nisbett; Dov Cohen

What example does the author give of the importance for developing some knowledge about the cultures under study?

Richard Shweder: wrote about a team of psychologists from Scandinavia studying variations in the "universal" family meal, India (local psychologist, uncomfortable, getting up from table), family meals should not be presumed to be part of some universal grid

What are some methods the author lists that are particular to the study of culture? These methods help us come closer to the idealized goal of being able to manipulate culture.

Situation sampling, cultural priming, culture-level measures

What are two methods that approximate experimental manipulations of cultural background?

Situation-sampling; cultural priming

What did Nisbett and Cohen find?

The South has a culture of honor that leads people (particularly White men) to respond aggressively to insults or threats to their honor, and this seems to be associated with the herding background of the South's initial European settlers. Herders face a particular kind of threat that farmers don't: their wealth is portable. Practice on marginal land that can't support large populations, making it very difficult to police. One would fare better if he/she could develop a reputation as someone who maintains his sense of honor (will respond with violence when people try to take advantage of you), and you'd want to develop this reputation beforehand.

Cultural priming

The activation of cultural ideas within participants. Making certain ideas more accessible to the participants. Since most of the psychological processes that have been studied thus far are at least existential universals, there are some ways of thinking that might be more uncommon in Culture A than Culture B. However, those ways of thinking are likely still present to a limited degree in Culture A as well. Therefore, when cultural ideas are activation that are more common in another culture, people start thinking in ways that are more similar tot he thinking of people from other cultures (American and Chinese participants: think either of how different from other (independent) or similar to family and friends (interdependent prime); describe themselves in an open-ended survey; self-descriptions became more similar to the other culture depending on what they were asked to think about).

Power

The capability of a study to accurately detect an effect (e.g., a cross-cultural difference) to the extent that one exists; a reflection of how well-designed a study is.

Generalizability

The degree to which research findings about the particular samples studied can be applied to larger or broader populations.

Occam's razor

The principle that any theory should make as few assumptions as possible; it maintains that, all else held equal, the simpler theory is more likely to be correct. The simpler theory is more likely to be correct. It's more parsimonious and more likely to be correct than four separate explanations.

What is the particular strength of Nisbett and Cohen's research program?

They used a wide array of different methods to test their hypothesis.

What is the advantage of the field experiment?

We can be confident we are observing patterns that generalize to the real world and are not just limited to the artificial confines of the laboratory.

Moderacy bias

When answering questions that require choosing a response from a scale (e.g., from 1 [Strongly Disagree] to 7 [Strongly Agree]), the tendency to choose a response near the midpoint of the scale. Remember: the respondents don't actually differ in their degree of impulsivity, just the way they respond. Independent of the content of the item.

Extremity bias

When answering questions that require choosing a response from a scale, the tendency to choose a response near the endpoints of the scale.

Remember: From colonial times to the present, the South has historically been more tolerant of corporal punishment of children, capital punishment, and gun ownership, and has been more supportive of the US engaging in wars compared with the North.

Y

Remember: Researchers often utilize methods commonly used in subfields of psychology because culture's influence on psychological processes cuts across almost all subfields of human psychology. However, this can create problems because there may be issues with the methods, and more issues may be created when being used to study cultures.

Y

Remember: The author writes, "My own personal view...is that we should be suspicious of any cultural differences that are identified by comparing means across subjective questionnaire measures unless the patterns converge with findings from other methodologies."

Y

Remember: no single study is perfect. It is good practice for the researcher to try to replicate a finding with a different kind of method. Multiple studies will give us a more complete picture from multiple perspectives. Multiple studies with divergent methods = if a different explanation is offered, it must pertain to all of these methods. Occam's razor pertains somewhat to this idea, maybe (simpler theory is more likely to be correct).

Y

What did Patricia Greenfield find when studying the _____, an indigenous people in _____?

Zinacantecans; Mexico; Her participants became angry when she repeated a similar/same question over and over again because they hadn't been exposed to this way of research (or maybe even to research at all). She was using the methodological technique used in survey research of asking participants about the same issue in a number of items that vary slightly. This reduces concerns with random error and ensures that people's responses pertain to the underlying construct.

Subjective self-report measures make it problematic to compare average scores ___ cultures, but these measures can be useful for identifying individual differences ____ cultures.

across; within (within-culture validity of subjective self-report measures is preserved; can be used to identify correlations between different constructs within cultures; cultural members tend to share the same response biases and reference groups)

What kind of music was found to be more popular among working-class Americans, and what kind of music was more commonly listened to by upper-middle-class Americans? Why?

country music; rock music; Country music lyrics convey more messages about resilience and rock music lyrics convey more messages about uniqueness, about carving one's own unique path.

The vast majority of cross-cultural research has been conducted between _______. By far, the most common comparisons are between _____ and _____.

industrialized societies; North Americans; East Asians

What are some safeguards to take with coding?

keep coders blind to the hypotheses; employ multiple coders

Moderacy and extremity biases are ____ because they affect how an individual responds to an item independent of the content of the item.

response styles