LB, Research methods, WEEK 2

Question 1

What is science?

Answer 1

• Popper says that science is a collection of theories which can predict the future and be falsified
• E.G. Popper argued Psychoanalysis was not scientific because it could not be falsified > Freud uses concepts which cannot be measured and are within the unconscious so cannot be accessed nor falsified
• Scientific things cannot make predictions which have been proved to be wrong (otherwise it is an assumption)
• Key scientific method is experimentation + observation

Question 2

Philosophy of science in Psychology

Answer 2

• Have been many debates overtime arguing over whether psych is scientific or not > Kant (1786) argues psychology is not scientific because we cannot actually measure how people think/feel + it cannot be quantified thus not a science. Wundt (1864) argues back that we can measure these things (make people reflect on own perception) using experiments + controlled settings
• Behaviourist perspectives in psych also support that psych can be scientific by looking at things like stimulus-response > behaviourist perspective emerged leading to a move away from Freudian views > this view is much more scientific
• There is a debate between, “we cannot have internal states + analyse them” while behaviourists argue we can change situations + study the different effects of this on behaviour

Question 3

History of Personality & Social Psychology

Answer 3

• First personality theorist: Sir Francis Galton who discusses the lexical hypothesis > this means you can take most important personality traits existing + reduce them down in to specific + stable personality traits (reduce words describing them down more into factors of their personality) > the most important personality traits in language, reduce down, get stable personality traits through factor analysis
• Lexical hypothesis is based on the assumption that the most important personality traits eventually become a part of their language
• First social theorist: Kurt Lewin who argued peoples behaviour was a function of personality + their environment > neither personality or EV predict all of behaviour, it is the interaction between the two > important because it doesn’t focus on the past but understanding the situation (dominant view of Freud) > this view is abbreviated as F(P.E)

Question 4

Present day of personality & social psych

Answer 4

-We have explicit + implicit assumptions when we do scientific studies
- Explicit assumptions: in differential psych, assumes that intrapersonal factors (within the individual) can be understood > understanding individual factors can help predict behaviour + describe future behaviour, factors of peoples personalities can be separated + verify introspection (self-reflection) through people’s behaviour
- Social psych indicates that intrapersonal effects are measurable, significant + have similar effects across people > social factors have similar impacts broadly. If they don’t, factors can be broken down + build a framework (e.g. take a breakdown of two people meeting for a data + find the commonalities between them + change one thing to see what happens such as wearing a fur suit vs not)
Implicit assumptions: The tools we use (e.g obs) can analyse information + we can construct an overarching theory of behaviour using these methods + tools

Question 5

Methods of data collection

Answer 5

• How we collect meaning what kinds of data we are gathering (e.g. objective measures as in watching what people do if we poke them with a stick or run experiments) > direct obs, interviews, self-report, experiments
• By time meaning how long the experiment goes on when collecting data > cross-sectional, longitudinal, experimental
• By sample meaning by case studies or meta analysis
• These all overlap

Question 6

How we collect data

Answer 6

• Direct obs: experimenter observing people’s behaviour/interactions (e.g. Gottman observed married couples conversing + predicted if they would divorce
• Interviews: e.g. asking someone w/ PTSD, how did you perceive a situation
• Self-report: e.g. 1-10 how do you feel about this
• Exp: manipulate something in one group
• Objective measures: assess brainwaves
  Subjective measures: How did you feel about this? Overlaps with interviews + self-report

Question 7

Macrolevel issues of studies

Answer 7

• If we do any test or any type of analysis, we have questions of sensitivity vs specificity
• Sensitivity: will you find what you are looking for if it is there (true + rate) while Specificity is: will you find what you are looking for or show something else (true - rate)
• E.G. If you scan someone’s chest to see if they have COVID-19, it has good sensitivity (can detect covid if it is there 93% of the time) but low specificity, won’t show something else(32% of the time someone has COVID, it may not be found + may be a positive test due to something else)
• Effect size: Is there a big enough impact/change? E.g. if we are measuring video games vs violence, the effect may be so tiny thus doesn’t do much
• Effect size changes our ability to detect things + distinguish between things
• Sample size: this is related to the above issues > if we have a big sample, we will often get some kind of information regardless of whether it is true (e.g. if we look at video games + aggression on a very large sample, it is likely to find a link by chance). But is this a real link? The effect size may be very small
• Depending on what we are studying, we will require a certain sample size > e.g. if we are looking at perception of dots moving, we don’t need a large sample as if ppts are normal sighted, the results should be similar (50-100 is fine). MRI scans only use 12-20 people as we assume our brains look the same
• But 50-100 ppt is not enough for social psych as people are too different + too much variance so a larger sample size is needed so we know if what we see is real + impacts the above

Question 8

Validity

Answer 8

-How do we know what we found is correct > more types of validity exist
- Basic types of validity includes internal vs external > internal: does it measure what it claims to + external: can it be applied externally to a situation
-Specific types of validity too like criterion, content, construct, discriminant
• Criterion validity: do the measures link to an observable outcome (current + future) essentially, criterion validity is predictive + concurrent validity combined (predictive = does the test predict later performance on a related criterion, concurrent = does the test relate to an existing measure) > basically compares the current test to existing valid measures (criteria) > links to sensitivity
• Construct validity: = to internal validity for things which are not operationally defined > does it measure what it says it will? Construct validity must be investigated whenever no criterion or universe of content is accepted as entirely adequate to define the quality to be measured.
• Content validity: measure all parts of the domain being measured (e.g. depression test must assess all sub-types of depression or else lacks content validity)
- Discriminant validity: Does it measure things which are different from the theoretically related things? Links to specificity (do we find things we didn’t intend to)

Question 9

Reliability

Answer 9

• AKA consistency > will you get the same result if someone else conducted the test for example
• Do you get the same result if you do test-retest, different person gives the test, you do different versions of the same test (e.g. reverse items on test) + check individual items relate to the overall thing
• Is needed for true validity, but does not mean it is valid alone if there is reliability
• Can have good reliability but poor validity, poor reliability + validity or good reliability + validity > reliability is needed for validity essentially
• Reliability is often expressed as Cronbach Alpha which refers to how closely related the items are (e.g. I like sweet things, I like sugary things > cronbach alpha shows degree of similarity within people)

Question 10

Reliability: Factor Analysis

Answer 10

• Reliability wants to be able to achieve the same result consistently to show it is reliable. But if there are questions which are not reliable (not matching up), factor analysis can be used
• Factor analysis looks at patterns of intercorrelations between the question responses, if they have the same/similar correlation, that means you are assessing the same thing > this enables us to come out with patterns between groups + variables
• These patterns will give a “loading” against a theoretical “factor” (something hidden being assessed + looks at dimensions across this point) > this theoretical factor is theoretically important (would expect it to be important)
• FA is used to extract latent variables (unobservable) which may explain correlations between different variables which you are able to measure > when FA finds certain patterns, this can be used to extract latent variables (because they are latent + not directly measured, they are more like theoretical constructs) which can support these theoretical variables

Question 11

Factor analysis of previous scale

Answer 11

• One type of factor analysis is principle components analysis > takes the biggest factor + takes out maximum variance against second factor + so on > if there is a latent variable within a set looked at then another variable should be considered to look at too
• E.G. if there are 30 questions on a socio-path scale + use PCA, we use it by looking at the Eigenvalues, if this is above 1, then we consider it to be a true factor
• When we run PCA, we find everything loads into one factor or we get two or more factors involved too
• Limitation: doesn’t encompass all variation which may exist

Question 12

Would this sociopath scale be valid?

Answer 12

• First need to ask is it reliable? Yes, if we test people over and over, we get the same result > people’s food taste is stable
• Is it valid? No, it fails criterion, construct and content validity > aren’t measuring what we intend to measure + doesn’t measure all domains of being a sociopath
• Many measures/scales existing already have this problem: e.g. Rorschach test has poor internal validity + Hamilton depression scale has poor content validity
• Validity is why theory matters > theories on depression for example inform content validity > we know to include certain things to cover all components

Question 13

Debate between social & individual differences

Answer 13

• Differential psych tends to have very reliable scales but sometimes have poor validity > personality tests which we currently use have poor validity because it doesn’t encompass all types of personality + the factors we use to distinguish between personality traits may overlap
• Social psych tends to have scales which are not reliable, people tend to change their response on different kinds of scales quickly but there is strong validity of concepts in some theories (what is measured is true)

Question 14

Understanding the split in Social Psych

Answer 14

• Exp + critical psych diverge at certain points due to core assumptions about the types of approaches which can be used to understand the person
• Experimental/traditional approaches prioritise scientific methods as the valid way of understanding the social world as it is the assumption that we can understand parts of human behaviour through identifying underpinning factors which help us conceptualise + measure human behaviour
• Critical approaches alternatively prioritise qualitative methods as understanding the social world > looks at level of talk + what this can tell us about the person + the society they are in
  This divergence come from historical differences in Psychology

Question 15

Roots of Qualitative Methods in Social Psychology

Answer 15

• Questions which we now see as psychological were initially discussed by philosophers > philosophies, especially on nature of knowledge + reality, underpin our current understanding of social psychology
• Essentially, philosophy helps us understand the person + helps us understand how we can come to measure the persons behaviour
• William James (1890) has been seen as the father of Psychology + wrote on the Principles of Social psychology > wrote in 2 volumes where one focused qual methods like talk + interaction while other volume discussed more traditional + empirical methods
• This work came from a critical view of introspection, saying it did not address the “connectedness” of human thought > essentially documents there are different ways of understanding human experience + shouldn’t necessarily be done apart (likes mixed methods)
• Völkerpsychologie (1900s) was an early movement in social psych in Germany + translates to “folk psychology” but better thought of as “psychology of the people” > seeks to understand how everyday people make sense of their social + personal worlds
• Volkerpsychologie focuses on the link between culture + language > this underpins current critical social psych > e.g. one of the founders, Lazarus + Steinthal looked at how language could be used to examine social knowledge (lan + culture are intertwined + important in coming to understand our social world)
• Wundt (father of exp psych) wrote on volkerpsychologie (looking at the intersection between language + social knowledge) then shifted to focus on exp psychology

Question 16

Roots: Divergence

Answer 16

• There was a split in approaches, philosophy + the role of the person and society > European + American split
• America developed it’s individualistic culture (social forces determined by the person) > Psychological Social Psychology (focuses on individuals + social factors impacting behaviour)
• European philosophy + psych focused on collective + behaviour influenced by groups (culture) and organisations > Sociological Social Psychology
• PSP may represent more traditional + experimental social psychology while SSP focuses more on critical social psychology
• Roots of experimental psych involve development of focus on individualism, behaviourism, cognitive psych, social learning and gestalt psych (context is important in perception)

Question 17

Roots: Turn back to Language

Answer 17

• Reconnected to Europe + Sociological social psychology > happened due to response in criticism to quantitative + experimental methods > start of critical social psychology
• E.G. Moscovici (1960, 1970) was known for leading his town back to language > developed social representation theory (looked at social representations of “madness” in everyday people + interested in how language of “experts” laid understanding of given phenomena > found people would use this language to make sense of behaviour of others
• Language at the time provides context + gives meaning to experiences that people take up and use
• Critical social psych can be interpreted in many ways leaning into different areas (more sociological or new paradigm in social psychological) > importantly they are all underpinned by use of qualitative methods
• E.G. social constructionism by Gergen & Gergen saw critical social psych as a new paradigm within social psych > critique of taken for granted knowledge underpins social constructionism (have to be sceptical of prevailing knowledge + understanding > why do these ways of thinking exist)

Question 18

Data collection in Critical Social Psych

Answer 18

• This focus is not on using experiments to gain knowledge + understanding, it wants to explore language to understand how people talked about their experiences + phenomena
• Ethnography can be done covert or overt
• Focus on seeking understanding and insight, rather than cause and effect, no need to turn constructs into variables, measure factors, aim for control, and exclude those that cannot be controlled
  No claim to objectivity, representativeness, realism, or control > Instead interested in exploring ‘talk’ (this can come in many forms)
• Asking appropriate questions, with appropriate data collection and analytic methods, conducted rigorously and ethically

Question 19

Ethnography

Answer 19

• Embedding yourself in the community of which you’re studying. Can produce audio, field notes, and photographs
• e.g Bengry-Howell & Griffin (2007): ethnographic approach to understanding masculine identity construction of working-class men through car modification

Question 20

Interviews

Answer 20

• A one-to-one purposeful discussion exploring the research topic. Can be conducted in-person, online, on the phone. Usually produces audio, but can also include video and images provided by the participants (photo-elicitation).
• Pickens & Braun (2012): interview study looking at young adult heterosexual women’s experiences of being single

Question 21

Focus Groups

Answer 21

• A purposeful group conversation, similar to focus groups, but also addresses interactional aspects and explore concepts in more depth.
• Griffin et al. (2009): focus group study exploring young people’s narratives around drinking and loss of consciousness/memory

Question 22

Diaries

Answer 22

• Participants encouraged to write and document parts of their lives, this can be done based on specific intervals, or more reactionary to events/emotions. Usually written, but can be done through an app and include pictures/video/audio.
• e.g. Boserman (2009): diary method to explore the meanings behind cannabis use.

Question 23

Documents

Answer 23

• These are usually pre-existing documents, analysis focussed on key messages in this data source. Can include newspapers, guidelines, or advertisements, for example.
• e.g. Medvedyuk et al. (2018): critical review of obesity and health literature to explore the constructed relationship between health and weight.

Question 24

Internet

Answer 24

• These can be created for the purposes of research or existent content can be analysed (though ethical issues in this). Can be blogs, chat rooms, websites, message board, social media sites.
• e.g. Fellenor et al. (2018): explored the social amplification of risk about ash dieback disease on twitter,

Question 25

Story Completion

Answer 25

- A type of projective test wherein participants complete a story stem. In so doing, it is hoped to ‘tap into’ ways of thinking and override barriers of admission. Useful for looking at a range of assumptions of a given topic. - e.g. Tischner (2019): used story completion to explore constructions of weight-loss motivations and health stories.

Question 26

Data analysis in Critical Social Psych

Answer 26

- As the focus is on exploring meaning and understandings, quantitative analysis of the textual data is not enough-it will only show frequencies/maps of specific language use - Qual analysis aims to discuss patterns of meaning in the text-this needs to go beyond description and demonstrate interpretative insights The type of interpretative insights are different depending on which analytic technique is used > different analytic techniques help answer different research questions

Question 27

Thematic analysis

Answer 27

- One of the most common qualitative analyses, seen seen to be a cornerstone of other qualitative analyses. TA tends to focus on the identification of common themes, and is good for exploring topics.  Not tied to epistemological positions. - e.g. Clarke & Braun (2020): used TA to analyse story completion data exploring stories of male body hair depilation

Question 28

Conversation analysis

Answer 28

- How interactions are represented via fine-grain analysis talk and what action the talk represents in (usually) naturally occurring conversations, e.g. the role of turn-taking and pauses. Focus is on the process of interaction  - how it is managed and constructed - e.g. Kitzinger & Frith (1999): Conversational analysis of focus groups exploring sexual refusal

Question 29

Discourse analysis

Answer 29

- Talk exists as a function of social action, and that language is used purposefully. Discourses represent patterns in the data that reflect something of the constructed nature of the phenomenon, with the focus on language as central to how the phenomenon/population is conceptualised. Heavily informed by social constructionism, but researchers can take different positions/approaches/theories, such as Foucauldian-informed DA - e.g. Riley et al. (2010): use of FiDA to explore neo-liberal constructions of magic mushroom use from focus group data

Question 30

Interpretive phenomenological analysis

Answer 30

- IPA focusses on individual sense-making and focusses on experiential knowledge. It is strongly associated with critical realism, phenomenology, and symbolic interactionism. - e.g. Van Parys et al. (2014): IPA study examining retrospective accounts of adults who had a mother with depression using interviews

Question 31

Narrative analysis

Answer 31

- Based on the assumption that people story their own worlds and experiences, researchers take a narrative approach to interviewing. There is an attempt to interpret the story as a whole, rather than in themes or discourses. Different approaches within NA. - e.g. Fisher et al. (2018): explored the narrative experiences of women in a mix-gendered gym

Question 32

Assessing quality in qualitative research: sensitivity

Answer 32

- When reporting on experiences, it should be embedded in the context of peoples talk - Appropriate language - Literature - Ppt perspective + socio-cultural context - Ethical issues - Taking an inductive approach to data interpretation

Question 33

Assessing quality in qualitative research: commitment and rigour

Answer 33

``` Remaining faithful to the participants’ stories • Appropriate data collection • Depth/breadth of analysis • Methodological competence & awareness • Engagement with the topic ```

Question 34

Assessing quality in qualitative research: transparency

Answer 34

Documenting & demonstrating interpretation - Retain & outline analytical process - Use appropriate quotes - Refer back to the research question - Methodological transparency - Reflexivity

Question 35

Assessing quality in qualitative research: impact

Answer 35

- Reaching influence > impact of research beyond the confines of the study - Applied settings; does it inform or develop current practices? - Theoretical considerations; does it advance theoretical approaches? - Can it be used politically to help that population?