Methodology Flashcards
1
Q
how do we decide which cultures to sample?
A
- If goal is to determine whether some cultural variable shapes some aspect of psychology
- Pick 2 cultures that differ only in that variable (one is high, one is low)
- Not true experiments - cannot make strong causal claims, but shows us that they’re related
- If the goal is to see universality of some aspect of psychology:
- Pick 2 cultures that are maximally different
- If you find that these cultures both exhibit the variable of interest, you have some evidence for the universality of that variable
2
Q
3 ways to generate ideas
A
- Deduction: going from theory → hypothesis → prediction (general to specific; top-down)
- Induction: going from observation → hypothesis → theory/general expectation (specific to general)
- Abduction: going from 1 specific observation → figuring out how to explain it (starting with specific, figuring out which pre-established general expectation matches it; not trying to create theory)
3
Q
theory vs. hypothesis vs. prediction
A
- overarching framework that organises and explains phenomena and data; generate hypotheses that you can test to discover the boundaries of the theory (most general)
- hypothesis: a tentative statement about about a relationship that may or may not be true
- prediction: specific statement regarding the expected outcome of the study you’re running (most specific)
4
Q
2 primary research designs (compare them)
A
- Quantitative vs. Qualitative:
- Type of data: qualitative = text, narrative, personal experience; quantitative = numerical data (ex. Likert scales)
- Goal: qualitative = to describe experience; quantitative = to identify social regularities
- Type of research: qualitative = typically inductive; quantitative research = typically deductive
- Methods: quali = unstructured (minimal prompting) or semi-structured; quanti = structured
- Sample size: quali = small; quanti = large
- Information per participant: quali = large quantity; quanti = variable quantity
- Type of analysis: quali = subjective, interpretive; quanti = statistical, summarization
- Ability to replicate: quali = low; quanti = high
- Generalizability: quali = low; quanti = high
5
Q
5 quantitative methodologies
A
- Explicit measures: asking participants to directly report their thoughts and emotions (ex. Self-report questions)
– Psychological surveys usually done using number scales (Likert-type scales) → good for comparing within groups, but can be problematic comparing between groups - Implicit measures: measures for attitudes over which participants have no conscious control (ex. Implicit Associations Test → you match things faster if they’re associated in your own mind - ex. black bad vs. good)
Behavioural measures: actual behaviours related to conceptual dependent variable (ex. Amount of food eaten)
Neurological measures: use neuroimaging techniques to determine neurological changes and role of neural structures (ex. fMRI)
Physiological measures: body’s automatic reactions to stimuli, excluding changes in the brain (ex. Galvanic skin response)
6
Q
issues with psychological surveys
A
- response biases: systematic tendency to respond to questionnaire items on some basis other than the specific item content → threatens validity of cross-cultural responses
- reference group effect: how people respond to questions depends on the group they’re using for reference
- deprivation effects: when there is less of something in your society, you value it more
7
Q
solutions to reference group effect
A
- use objective measures rather than subjective → ex. Rather than asking “are you tall?”, ask “how tall are you?”
- use specific context-based behaviours or scenarios rather than abstract psychological concepts (ex. emotions, traits) → ex. Rather than asking “Are you conservative?”, ask “Do you believe it’s wrong to engage in premarital sex?”
- Using behavioural measures (ex. Measuring pace of life by watching how quickly people walk) or physiological measures
8
Q
deprivation effects
A
- when there is less of something in your society, you value it more (ex. If your culture has lower personal safety, you’ll value personal safety more than someone in a culture when they rarely have to think about it) → difficult to make inferences about cultures based on values people endorse
- No way to truly avoid, but helpful to cross-check your results about values with the results from other studies
9
Q
qualitative methods: important considerations
A
- Understanding the cultures you’re measuring
- Methodological equivalence (are you methods equivalent in the cultures you’re studying?)
10
Q
qualitative methods: research considerations
A
- Researchers must understand norms and practices of other cultural environments (especially regarding the psychological phenomenon in questions)
- Risk conclusions based on faulty information and assumptions
- Accomplished through ethnographies and international collaboration
- Taking research beyond the lab → field research
- Cultural psychologists/anthropologists go into the field to examine human psychology and collect data about research questions (ex. ethnographies)
- Field = anywhere not in the lab
- Indigenous peoples’ research
11
Q
ethnographies
A
Comprehensive collection of data and knowledge about a particular cultural group through interviews, cultural immersion, etc.
12
Q
indigenous peoples’ research
A
- methodologies make heavier use of qualitative methodologies than current mainstream academia (ex. Storytelling, conversations)
- Cultural practice of oral tradition → greater familiarity with expressing ideas qualitatively
- Because of mainstream psychology’s reliance on quantitative methodologies, indigenous knowledge receives little recognition (many psychologists are not trained in interpreting qualitative data and results) → makes it difficult for Indigenous researchers to thrive, which means that there aren’t enough people producing Indigenous knowledge (or what exists is ignored)
13
Q
Amalgamating methods
A
Forms of communication, created by people within a cultural environment, that reflect the overall psychology of the people in a cultural environment
14
Q
challenges in translating materials for research
A
- Certain nuances are lost
- Words/syntax may not carry over
- Have to decide between conveying exact sentence or general idea of the sentence
- Certain psychological concepts don’t carry over (ex. self-esteem)
15
Q
ways to translate research
A
- Have a bilingual collaborator that can decide whether translated materials are appropriate
- Challenge: perhaps that person’s understanding of the methods/words is idiosyncratic
- Back-translation: translator 1 translates from original language to target language → translator 2 translates materials back to original language → original and back-translated materials are compared and discrepancies are resolved
- Consensus method: multiple bilingual speakers arrive at consensus
16
Q
structure of a questionnaire
A
- Some concepts have “factors” in their structures
- structural/construct equivalence: structure underlying psychological construct is comparable across cultures
- Factor analysis: statistical analysis that determines how many factors can be separated in data → can use this to figure out the best solution (1-factor, 2-factor, etc.)
- – Analysis generates a “fit index” to determine which solution fits the data best
17
Q
structural/construct equivalence - once you have it, what do you do?
A
- Identify and keep questions that appear to be understood in similar ways across cultural environments
- Identify and eliminate questions that appear to seem to be understood differently cross-culturally
18
Q
issues with equivalence
A
- Researchers also need to watch for methodological equivalence:
- Ensure methods are understood in identical ways across cultures
- Some cultures may not understand process of completing psychological surveys
- Researchers may need to use slightly different methods with different cultures (especially dramatically different ones) → causes problems for making direct comparisons, but less problematic than asking participants to do things they don’t understand
19
Q
3 types of research designs
A
- experimental
- correlational
- quasi-experimental
20
Q
experimental designs
A
- Researcher manipulates the study conditions that participants will be placed in
- Each condition is a randomly assigned group
- Researcher has complete control over the IV, can determine cause/effect
21
Q
correlational design
A
- Researcher measures all (continuous) variables they’re interested in
- No “groups” (unless it’s also part-quasi-experimental and has 2 groups that have inherent order → more than 2 groups cannot calculate meaningful correlational coefficient)
- Researcher has no control over any of the variables, can only look at relationships
22
Q
quasi-experimental designs
A
- Researcher measures dependent variable as a function of groups
- Groups are naturally-occurring (ie. culture), no random assignment
- Researcher has no control over groups; can only look at relationships
- By nature, cultural studies are quasi-experimental designs
23
Q
2 kinds of IV manipulation
A
- Between-groups manipulation: different groups of participants receive different levels of the IV (require random assignment, classic experiment)
- Within-groups manipulation: each participant receives more than 1 level of the IV (no random assignment, participants experience all conditions); careful of order effects here
24
Q
types of response biases
A
- moderacy bias
- extremity bias
- acquiescence bias
- socially desirable responding bias
25
Q
moderacy bias & extremity bias
A
- moderacy: always choosing the middle/neutral item (aka: fence-sitting)
- More common amongst those with Asian cultural background, especially if questions seem controversial
- extremity: only picking the most extreme items (strongly agree/strongly disagree)
- More common amongst hispanic and African-American populations
26
Q
how to avoid moderacy and extremity biases
A
- doing yes/no questions or forced-choice questions
- Issues with these: eliminates subtle differences/nuances that might be really important
- Can also be avoided by standardizing scores so you can compare how much each one deviates from the participant’s personal average
27
Q
acquiescence bias (and prevention)
A
- always saying strongly agree (aka: yay-saying)
- More common amongst those with Asian backgrounds; common when people don’t fully understand the question
- This bias can be eliminated using reverse-scored items
28
Q
socially desirable responding bias (and prevention)
A
- answering in ways that make us look good
- to avoid:
- – Providing anonymity
- – Using neutral items to describe both positive and negative characteristics (ex. “I am comfortable talking to people I don’t know well” vs. “I am friendly”
- – Use separate measures to assess participants’ tendency to engage in socially desirable responding (ie. Socially desirable responding scales: socially desirable responding occurs due to two different motivations → self-deception and image management; example: Paulhus Deception Scale → true/false, items like “I never swear”, “I sometimes speed” etc.)
