Chp2 - Research Methods in Social Psychology

hindsight bias

• exaggerated ability to foresee results after learning the outcome

Menec & Weiner (2000) - example of hindsight bias

• interested in p’s opinions on genetic screening for disorders before deciding to have children

• p read statement → woman refused to test

  • either told woman later had a child with a genetic disorder or not

• P were asked to judge the probability of woman having a child with a genetic disorder

  • those who were told she would respond with knew-it-all-along

  • blamed the woman

consequences of knew-it-all-along phenomenon

• arrogance → overestimation of intellegence

• makes us more likely to blame the decision-makers for ‘obvious’ bad but praise them for good choices

• self-blame

approaches to doing research

• quantitative

• qualitative

differ in epistemology

epistemology

• the study of knowledge & the underlying status we give it

  • e.g fact or a version

Quantitative social psychology

• aimed at studying relationship between variables

  • expressed numerically

  • explored via statistical analyses

• see social psychological world in terms of variables

  • a thing that can vary in quantity/quality

• a variable is any conceivable characteristic e.g demographics, feelings, behaviour

  • can change overtime & vary low/high

• study interplay between variables

  • measure variables under scrutiny

Qualitative social psychology

• sceptical of simplifying complex thoughts & feelings into numbers

• numerical ratings dont capture depth of experience → more complex

• focuses on interpreting words & behaviours

  • use interviews or real-life observations

• believe human behaviour is influenced by context and culture → fixed rules don’t govern it

• P can reflect/react to psychological findings

  • can change behaviour = making it harder to predict

key differences for quantitative approach

• Pursue systematic measurement of phenomena

  • often in controlled laboratory settings

• Make predictions about the outcome of research

• Aim at establishing general laws & principles about types of phenomena

key differences for qualitative approach

• Focus on interpretation of phenomena as emerged in naturalistic, unconstrained situations

• open to new, surprising & previously unthought-of findings

• Aim at providing a thorough description & understanding of the specific phenomena under investigation

theory

• an interrelated set of principles that guide what should be studied, & explain/predict the observed relationship between variables

self-handicapping

• Blaming your equipment is a phenomenon

Correlational Research

• Focuses on studying natural relationships between variables.

• Example

  • Wagner et al (2015) longitudinal study examined self-esteem in 462 aged 70-103

  • found that self-esteem remains stable until the late 80s/90s, where it declines

  • Decline is linked to loneliness, poor health, disability, & less control

  • Those who stayed healthy, socially connected & felt in control had minimal self-esteem decline.

Longitudinal Study

• Research where P are repeatedly tested or observed over a period of time.

• Wagner study used both existing data and new data gathered through surveys and medical records.

Designing Questionnaires

• Questionnaires are used to assess relationships between variables (e.g., social status and health).

• Researchers ask questions related to the variables of interest (e.g., employment, health status).

• Questionnaires can be distributed via phone, online, or on paper.

5 important factors when designing a questionnaire

• nature of sampling

• order of questions

• reposone options

• wording of questions

• validity & reliability of measures

types of sampling strategies in correlational research

• Random sampling

• systematic sampling

• stratified sampling

• cluster sampling

• opportunity/convenience sampling

• snowball sampling

• Theoretical/principled/purposive sampling

Random sampling

Where everyone in the population under study has an equal chance of being represented in the sample

Systematic sampling

• Where members are drawn from a population at fixed intervals (e.g. every fifth person)

Stratified sampling

• Aims to ensure all features of a population are represented in the sample

Cluster sampling

• When the population is organized into groups (or clusters) and some ‘clusters’ feature in the sample

Opportunity/convenience sampling

• practical sampling method where the sample is made up of people who are easiest to reach or most convenient to include, usually because of limited time or resources.

Snowball sampling

• When the researcher ‘snowballs’ further participants from one respondent

  • (e.g. their family & friends)

Theoretical/principled/purposive sampling

• When P are chosen for inclusion in research on ‘principled’ reasons for their inclusion.

• Does not seek representativeness

Order of questions

• The order in which we ask questions may produce biased responses.

  • e.g people’s support for civil unions of gays rises if they are first asked their opinion of gay marriage (Moore, 2004)

• Therefore, we need to give careful consideration to the way we order the questions

Response options

• can greatly affect the answers you receive

• P will fit their answers into options provided → even if they don’t match.

• If the options are too limited → P answer might not accurately reflect their true response.

• It's important to provide enough to capture a wide range of possible answers

Wording of questions

• Exact wording of questions can influence answers

  • e.g P may support cutting "foreign aid" but favor increasing spending "to help hungry people in other nations,"

• How questions are asked in a questionnaire is important & requires careful consideration.

• Small changes in tone or wording can have significant effects on responses.

• Even if P feel strongly about an issue → way a question is worded may still influence answer

Validity & reliability of measures

• Questionnaire items are used to measure specific variables.

• For meaningful results → measures must be both valid & reliable.

• e.g. To measure academic self-esteem

  • "I regard myself as a competent student"

  • "I have good studying skills"

    • both accurately assess self-esteem & consistently work together.

Validity

• The items must accurately reflect the concept they are intended to measure (e.g. academic self-esteem).

Reliability

• The items should consistently measure the concept and produce the same results under similar conditions.

Association & Causation

• associations detects relationship

  • cannot show causation

correlation research

• study of naturally occuring relationship between variables

types of correlation

• negative correlation (-0.6)

  • 1 variable increases as the other decreases

• positive correlation (+0.8)

  • 2 variables increase or both decrease

confounding variable

• uncontrolled variables that interact with IV which affect outcome of research

• researcher can then determine which variable is responsible for observed results

randomly assigned conditions

• if p cannot be randomly assigned → correlational study

• if P can be randomly assigned → experimental

independent variable

• the experimental factors that a researcher manipulates

dependent variable

• the variable expected to be dependent on the manipulation or change in the independent variable(s)

experimental conditions

• condition in which the IV is presented to measure its effect on DV on P in this condition

control conditions

• the condition where the IV is absent and the data generated from this condition is used to compare with the experimental condition

random assignment

• process of assigning P to the conditions of an experiment such that all persons have the same chance of being in a given condition

Experimental Research

• Seeks to understand cause-effect relationships by manipulating one or more IV while controlling others

Quasi-Experiments

• Used when research can’t be conducted in a lab but still follows experimental principles.

• 2 basic types

  • natural experiments

  • field experiments

Natural Experiments

• researcher does not manipulate the IV, but instead observes natural differences (e.g. gender)

  • e.g Measuring time spent shopping by male vs. female shoppers without controlling gender

• Strengths

  • High ecological validity since it happens in real-life settings.

• Limitations

  • Lack of control over other factors (confounding variables) which can affect results (e.g., weekly vs. daily shopping, or chatting with others).

ecological validity

• The extent to which study findings reflect real-life situations

Confounding variables

• Uncontrolled factors that interact with the IV and affect the outcome

• making it hard to determine which factor caused the observed results

Field Experiments

• Researcher controls the IV but the study takes place in a real-world environment.

• e.g Testing whether women are more helpful than men by ‘dropping’ bags of shopping in front of people in a shopping center & observing who helps

• Strengths: High ecological validity due to the natural environment.

• Limitations: Less control over other variables (e.g., a shopper’s ability to help may depend on an unseen factor like a back problem).

Generalizing from Laboratory to Life

• Social psych combines everyday experiences with experimental designs to deepen understanding.

• Everyday observations lead to experiments

• Results from lab often mirror findings in real-world settings

  • helping policy makers & platforms understand social issues.

• content of people’s thoughts & actions varies across cultures → processes behind them (e.g., how attitudes influence actions) tend to be similar.

Universal forces

• Despite surface differences, people across cultures are influenced by similar social forces

• showing that beneath our diversity, humans share fundamental similarities.

Replication

• essential in quantitative research to confirm that findings are reliable/repeatable across different times & places

• If an experiment reveals a truth about behaviour → should consistently show the same result when repeated

• Nosek et al. (2015) → trying to replicate 100 social psychology experiments, only 1/3 to half produced the same results.

Challenges to Replication

• Researchers are often in a rush to publish for career reasons → making them reluctant to replicate results.

• Journals typically prefer publishing studies with positive results, making replications (especially failed ones) less likely to be published.

Conceptual vs. Direct Replications

• conceptual replications → many published studies replicate previous using different methods or P = showing the generalizability of results

• Many researchers prefer direct replications (exact copies of studies) to confirm findings with certainty

• Christian Montag (2018) suggests more cross-cultural studies with international samples for globally valid results.

• Brian Nosek: To prove something, it should be tested the same way again (direct replication).

Concerns about Conceptual Replications

• Flexibility in reproduction methods can increase the chance of confirming a phenomenon even if it’s not real (Simmons et al., 2011).

epistemology

• the study of knowledge and the underlying status we give it. For example, is the knowledge we obtain about human behaviour a ’fact’ or a ’version’?

reflexivity

• to recognize the role of the researcher in the production of the research findings

interviewing

• a strategy to obtain qualitative data based on talking with and asking questions of research participants

• questions may be asked either directly or indirectly, and responses are given in an open format

different types of interviewing

• structured

• semi-structured

• conversational/unstructured

Structured interview

• Fixed questions asked in the same order

• Typically answered using a predefined set of response options

• Because response options tend to correspond to quantities, this type of interview is more commonly used in a quantitative approach

• questions may not reflect participants’ experiences or understandings

Semi-structured interview

• Contains key questions to maintain relevance

• Flexible order in which questions are phrased & presented during to suit the experiences of P

• Can build good rapport with interviewee

• Useful for studying sensitive topics and issues

• lacks reliability

• poor researcher control

Conversational/unstructured

• Contains key topic/s

• Very flexible wording & presentation of questions

• P driven to capture their experience & understanding of the phenomena under investigation

• Can build good rapport with interviewee

• Useful for studying sensitive topics & issues

• poor reliability → little control over the interview

focus group

• a strategy to obtain qualitative data based on a small group discussion about the issue of interest facilitated by the researcher

participant observation

• a research strategy in which the researcher spends time in close contact with the people studied (tribe, group, community, team) for a prolonged period of time in order to gain a deep understanding of their perspectives/practices

Grounded theory (Glaser & Strauss, 1967)

• challenges grand theories that imposed preconceived frameworks on data collection & analysis

• Core Principle of Grounded Theory:

  • data collection then hypothesis

  • Concepts are developed from the data

    • categories are formed from these concepts

  • Constant comparison helps to identify similarities & differences between categories, forming the basis of theory development

• doesn’t aim to find the absolute "truth" → seeks to explain the data without generalizing it into a larger "grand theory."

Discourse Analysis (DA) (Potter & Wetherell)

• Focuses on the idea that language constructs reality

• Sources of Data→ Interviews, focus groups, natural conversations, diaries, online forums

• evaluates responses in context before drawing conclusions

  • Uses interpretative repertoires → linguistic resources P use to understand their social world

• Critical Discourse Analysis → some argue that language is shaped by power relations in society

  • Power structures influence what can(not) be said (e.g. stigmatization of LGBT)

• used to study prejudice, racism, other social phenomena

Interpretative Phenomenological Analysis (IPA)

• Based on phenomenology → explores the relationship between mind & outside world

• Focuses on subjective conscious experiences & how people perceive/interpret their lived experiences.

• Founded by Smith → uses an idiographic approach = looks at individual experiences in detail

• uses data from interviews, focus groups, and written texts

  • To understand the P’s lifeworld

• looks for themes in the P’s descriptions of their experiences.

  • Themes are organized hierarchically & compared across different P to identify common experiences

Researcher ethics

• BPS code of ethics

  • respect

  • competence

  • responsibility

  • integrity

respect

• privacy & confidentiality

• respect

• communities and their shared values

• impacts on broader environments

• power issues

• consent

• self-determination

• compassionate care

competence

• possession of appropriate skills & care

• recognize limits of competence and refer to another professional when necessary

• advances in evidence base

• maintain technical and practical skills

• matters of professional ethics and decision-making

• take mitigating actions when required

• caution in making knowledge-claims

integrity

• honesty, openness and candour

• accurate, unbiased representation

• fairness

• avoidance of exploitation and conflicts of interest

• maintain personal and professional boundaries

• address misconduct

responsibility

• professional accountability

• responsible use of knowledge and skills

• respect for welfare of humans, non-humans and living world

• potentially competitive duties

informed consent

• requiring that research P be told enough to enable the to choose whether they wish to participate

deception

• only if essential and justified

IS SOCIAL PSYCHOLOGY SIMPLY COMMON SENSE?

• Social psychology is criticized for being trivial because it documents things that seem obvious.

• Systematic research methods reveal that ‘outcomes’ are more ‘obvious’ after the facts are known.