RMC
- Created by: jxw145
- Created on: 02-10-22 11:05
Research cycle
Without research the disciplines we know and recognise as Psychology and Neuroscience today would not exist. Our understanding of the mind, brain and behaviour has been established through a continuous cycle of theorization and investigation and it continued to progress via this cycle.
We can theorize to our hearts are content, but without research, it is just a theory with no scientific basis. It is based solely on educated guesses and speculation.
Errors in research project
When carrying out a research project, things can go wrong at different stages:
-Is hypothesis worded in way that's consistent with rationale and allows you to test what you want to test? How does this influence your methodology?
-Have you chosen best methodology for question you're asking? Is data collected in the necessary format for hypothesis and analyses?
-Does your interpretation of the analyses allow you to talk about what it means for your hypothesis?
-Have you contradicted a theory, and which is more likely wrong?
Hypothesis testing
-The purpose of hypothesis testing is to check if we are wrong about something going on (is nothing actually going on)
-p value tells us how likely it is that something MIGHT be going on (not how big/important that something is- effect size or not even that there is something)
-A low p value means there is a slim chance we are wrong about nothing going on BUT this does not mean that it is likely something is going on
-When reporting, you don't need to mention the null hypothesis and your interpretation should be cautious- you can never prove something is right with 1 study (or even from a collection of them- what you can do is provide support for the theory)
-however, 1 study can prove a theory wrong and we should learn from this
-in a report when you expected a difference don't say we had support that there was no difference- actually if there was a non-significant effect, it just tells you that there is no evidence for a difference
Rationales for research
-an explanation/justification for doing something - give the reasons for why your particular study was carried out
-needs to state what the research aims to achieve (purpose)
-presented in the introduction
-links past research (lit review) to the research you’re going to write up in your report
types of rationale:
-can be motivated by methodological problems in previous research
-can consider different theories’ explanations of data and whether there were any gaps or problems in these explanations
why is the research important?
-To extend prev research on effects of variable A on dependent variable B
-help to determine the reliability of a previous finding- replicate with new set of ppants
-address a prev methodological problem
-help to clarify inconsistencies in literature
-resolve theoretical questions or dev better theoretical models in your area
-compare two rival theories to see which gives a better explanation of the data
-be predominately exploratory in nature (didn't find much lit in this topic)
conceptual vs technical/operationalised hypotheses
Conceptual= describing an idea clearly and concisely using accessible language
Technical= describing an idea precisely and with reference to appropriate technical terminology
*When communicating in science, we need to be able to translate from one form of info to the other
*In communicating about research, we need to be able to summarise ideas conceptually, but also present info in technical/operationalised terns in order to provide info about how a study translated the concepts into measures and manipulations.
research question/hypothesis
Research question/hypothesis should be at the end of introduction:
-How do the insights from reviewed lit come together and what is expected to happen based on what existing findings?
-what specifically did your study test? (use conceptual terms then operationalise to provide transition and focus for the method section)
-in quantitative report we often see both a conceptual summary and operationalised statement
choosing your design
-it is very simple to compare effects of one IV only with two conditions (either between or within)
-there are also one-way designs- 1 IV only but 3 or more conditions in one experiment -there are also factorial designs- more than 1 IV, equivalent to 2 experiments simultaneously
-the interaction means there are added effects which allows us to examine how diff variables might work together
-this is taken further when there are more conditions in 1 or both of the factors
more conditions
2/3 dimemsions
interactions
-the number of interactions depends on the number of factors (each additional factor adds more possible combinations)
-n of factors/IVs determines n of interactions but n of levels/conditions does not
-life is complex- not just simple comparisons
-we can explore interactions
Why do we need ethical principles?
-research impacts ppl, policies and organizations
-adherence to ethical principles is essential for:
- the continuation of the discipline- ppl won't participate in our research if they don't trust or respect the discipline
- the credibility of the research- ppl will not believe our findings or recommendations if our beh has given them reason not to
- our credibility and safety- if you violate ethical principles, you can be fired, unemployable, lose funding, face criminal charges
- ethical approval processed and committees are a safety net- if they say no to your ideas, it is to protect you, your ppants and others.
Guiding principle 1
“Respect for the autonomy, privacy and dignity of individuals, groups and communities”
-Valid consent (this is not always straight forward!)
- is someone able to give consent (minors, ppl with learning difficulties, degenerative conditions, animals)- have to be an expert to know if someone can give consent
- even if you have 3rd party consent, the actual ppants needs have to be met + no harm.
- have distress protocol
-Confidentiality (links to GDPR here – data must be stored securely)
- our personal data is our property
- only collect demographic info you need
- ppants should know who has access to data
- there are exceptions ppants should be made aware of (duty of care or required by law)
Guiding principle 1 (2)
-Right to anonymity (this is not always straight forward either!)
- we assign ppant numbers/use pseudonyms and remove identifiable features (however sometimes, it is still possible to identify ppants- especially if the pop is small or if you have used certain sampling (snowball)
-Fair treatment (respect the participants views, identity, culture, choices etc)
- can disagree with ppants privately but not publicly
- being blind to ppants individual differences and culture is disrespectful and harmful
-Due process (avoid unfair, prejudiced or discriminatory practices)
- required to avoid unfair, prejudices and discriminatory processes
- you should examine your own biases
-A special note on autonomy: Participants/potential participants do not owe you anything; not even their time or data (right to withdraw). -You need to put the effort in to find participants.
Guiding principle 2
Scientific integrity refers to a commitment to high quality research at all stages (from design to completion/dissemination).
-Not adhering to this principle wastes everyone’s time, including your own.
-Unreliable findings/recommendations can lead to harmful/negative outcomes.
-The value of your work and your professional reputation/credibility will be seriously undermined if you produce and publish poor quality research.
-No study is ever perfect, but have you done the very best you could with what you had, and have you acknowledged the strengths/limitations?
Guiding principle 3
social responsibility- our work can (and should often) have real-world impacts and we must be aware of & prepared for potential consequences.
-Research should be for the common good.
-We have a shared collective duty towards the wellbeing of humans (and non-humans if you are conducting research with animals).
-We need to be able to work with others including people in other disciplines, relevant organisations, ‘gate-keepers’ and stakeholders.
-We must work within our competencies (e.g., do NOT give advice you are not qualified to give and never put yourself in situations beyond your competency).
-We and your future employers also have a responsibility to keep you safe while you are conducting research (risk assessment & mitigation)
Guiding principle 4
maximising benefit and minimising harm
-Maximising benefit: We always aim to maximise the positive impacts of our research for stakeholders and communities.
-We have a responsibility to prevent harm (e.g., to the discipline, our reputation, participants, groups, communities & society).
-We must also try to prevent potential misuse of our findings, techniques etc. Always be clear to avoid your findings being misunderstood, misconstrued or misused.
-We should always consider our research from the perspective of all of the people involved.
In practice
Second year students have asked us to approve their attitude scale investigating “Undergraduate student attitudes towards Covid-19”. They have two weeks left complete the project. Their proposed questionnaire includes 30 7-point Likert scale items such as “I have been seriously affected by Covid-19” and “My life has completely changed due to Covid-19”. Each person in the group will give the questionnaire to 10 students to complete including some friends. Participants will choose a random participant number to ensure anonymity and they will reassure their participants that no one else will see their answers. They will also allow participants to withdraw their data up to two weeks after participating.
Ethical approval was not given for this project for lots of reasons, e.g. -This is a very sensitive topic which is still fresh and on-going. -There was a genuine risk of harm to participants. For example, what if this questionnaire was given to someone who was grieving? -Data were not anonymised. The researchers knew the participants! -There was no opportunity for distressed participants to talk. Sometimes questionnaires are simply not appropriate for the topic. -The researchers did not have the competencies or experience to deal with the potential implications or consequences. -The items in the questionnaire did not assess attitudes!
*Sometimes, if topics are too sensitive, rather than ask about personal experience or behaviors, you can ask about the persons attitudes toward the topic.
ANOVA
ANOVA=Analysis of variance
VARIANCE= measure of spread/dispersion (in statistical terms, variance= (standard deviation) squared
What does ANOVA test?
Are 2/more groups from the same population of scores?
-in other words, are the differences in individual scores in each of the subsets of our data of a similar size? and are the differences across the whole dataset all of the same size too?
-in other words, are the groups equally as varied as each other in comparison? (g1=g2=g3)
*If yes, then we have no evidence of an effect
error variance
-Data sets can be represented by a bell curve- the peak of the bell curve represents the mean which is in the middle and the width of the bell curve represents the variance
-In a between ppants design with 3 conditions (e.g., laptop notes, laptop notes with review, and handwritten notes), we would expect our ppants within the condition 1 group to perform differently to each other due to naturally occurring differences within any group because ppants will respond differently
-this is called within group error variance or sometimes just ‘error’
-essentially error is a statistical term for the diff in responding due to naturally occurring variation in the absence of any effects
Bell curves
-we could represent the spread of data within condition 1 using a bell curve to represent the error variance (and for all the other conditions) e.g.
*Here the variance within each group is still comparable as the width of the curves are similar, but the peak of the curve is in a different place along the x axis in condition 3- ppants on average performed more highly than in 1 and 2
-variation between the groups
Difference between groups
The differences between the groups are made up of two distinct sources of variance:
- error- natural variation between people (between group variance as ppl are just as diff across groups as they are within)
- any effect of the factor we are studying (if we represent the between groups variances as a curve, this curve will be more spread out than the error only curves that we’ve looked at so far)
Between groups bell curve
*This variance takes into account the lowest scores from conditions 1 and 2, and the highest scores in condition 3
SO, across all 3 conditions, there is greater variance than there is within each condition
-and that’s basically what ANOVA works with:
*If the overall variance for the study is bigger than the error we would expect, then the variation we see is likely to be due to the effect of the factor that we are investigating
Within ppants design
(Same ppl in each condition- acting as their own control)
What does this mean in terms of sources of variance and error (the naturally occurring variation in ANOVA)?
-for the within group error variance, this will be the same as between ppants designs
-similarly, the effect of the factor we are studying is going to be much the same as the ppants in the between groups design
-what is different is the between group error variances (the ppants will be more similar to themselves on different occasions than they are likely to be to a diff person)
Within groups bell curve
*Should be represented by bell curves that are much narrower (error between conditions is smaller)
*This makes the within-ppants analysis of variance a more sensitive test so it is able to detect smaller effects than the between ppants ANOVA can
Summary
ANOVA compares naturally occurring variation called error to the overall variation across groups and that's called the between groups variance
-And if the variance that we observe within a group performing in the same context is the same as either the variance between different groups or the variance between different contexts for the same group, we conclude that there is no effect of the factor-
*We would say the results came from the same population
-If the variance between groups or conditions is greater than the variance within each group, we might have an effect
Visualising variance comparisons
*The between groups variance is greater than the within groups variance which suggests the 2 groups (g1 and g2) are sig diff to each other
-In statistical terms, the comparison between between-groups variance and within-group variance is captured in the F ratio
The F-ratio
-In statistical terms, the comparison between between-groups variance and within-group variance is captured in the F ratio
*This is a test statistic for each effect in ANOVA (in 1 way ANOVA there is one effect) -An F ratio that is smaller than 1 means there is no sig difference
-Essentially the F ratio tests the likelihood of the variance estimates being from the same population
-SPSS will provide an associated p value and provide an estimate of effect size- even more useful than p value -effect size explains how much of the variance observed in the data is due to the effect of the IV
-the more the variance is explained, the more important the effect is in scientific terms (bigger effect sizes are scientifically more important)
What does sig ANOVA tell us?
*That somewhere in the analysis at least one group is different from at least one other group…but where?
-It isn't possible to interpret ANOVA fully without considering the descriptive statistics associated with the analysis
-we need to look at which condition has highest/lowest mean and whether the differences between the conditions match the hypothesis we tested
-however, even if the descriptives appear to support hypothesis, we still have to check what differences between the pairs of conditions are sig
SO have to carry out post-hoc tests/follow up test
A post hoc comparison is a statistical test that allows us to identify which groups in an ANOVA are all significantly different to each other.
Post hoc test
-One way of doing this is to compare pairs of conditions.
(However, there are problems with multiple comparisons- familywise error rate)
*A p value measures the probability of our conclusion being wrong (making type 1 error)- we want the chance of making type 1 errors to be as small as possible
*In essence, the more tests we carry out on the same dataset, the greater the chance of making a type I error- for each comparison, the chance of drawing wrong conclusion about our data increases by the level of probability that we set as a criterion for significance (0.05)
Bonferroni correction
*So can’t really now claim our findings are sig
how to avoid family wise error- set an acceptable overall error rate (divide 0.05 by number of comparisons= Bonferroni correction)
-makes it a strict test
Non-parametric analyses
There are three main prerequisites for analysis of variance:
- interval/ratio data
- scores normally distributed
- homogeneity of variances
*ANOVA is fairly robust so can deal with data that is a bit skewed (e.g., response data)
-ideally also need a min of 8 data points per cell so there is variation in the data
Non-parametric tests
-Non-parametric tests are distribution free, use ordinal (ranked) data, no means of variances BUT lack sensitivity of parametric tests so greater chance of type 2 error
-when reporting non-parametric tests, you must state the assumptions for parametric tests that were violated, report the medians, min/max scores, and the required test statistics
-Ranking is a key part of non-parametric data analysis.
Types of non-parametric tests
The Kruskal-Wallis test is the equivalent to one way between participants ANOVA and
the Friedman test is the equivalent to one-way within participants or repeated measures ANOVA.
KRUSKAL-WALLIS TEST
- compares 3 or more independent groups
- similar to Mann Whitney U test
- uses ranked data
- Tests for differences between 3 or more treatment conditions, using a separate group of ppants for every treatment
-It ranks all scores across groups
If a real difference- the ranks for 1 condition will be systematically higher or lower than another condition
If no difference- the ranks for each sample will be mingled together
The Kruskal-Wallis test
The Kruskal-Wallis test tests the following hypothesis:
The ranks in one condition are systematically higher or lower than the ranks in another condition- there are differences between conditions.
As with an ANOVA, the test only says that there is a significant difference, but not where that difference lies and that means we need to follow up a significant Kruskal-Wallis test with post hoc comparisons.
*The test that we use for the post hocs is the Mann-Whitney test, which compares two independent groups but as with the ANOVA, we would also use a correction, a Bonferroni correction, to determine the criterion for significance for the Kruskal-Wallis test.
The Freidmans test
- non-parametric repeated measures one way ANOVA
- compares 3 or more related conditions
- similar to Wilcoxon- uses ranked differences
- tests for differences between 3 or more treatment conditions collecting data from the same group of ppants for each condition
The Friedman test tests the following hypothesis:
The ranks of the differences between a pair of conditions are systematically higher or lower than the ranks of the differences between another pair of conditions. As with ANOVA, we would again need to follow up a significant Friedman test with posthoc comparisons, and we would use several Wilcoxon tests and apply the Bonferroni correction to the criterion for significance.
Factorial ANOVA
ANOVA can be:
- Between participants (e.g., in 2 factor ANOVA would be 4 groups of ppl taking part)
- Within participants (e.g., 4 data points for each ppant- one for each combination of your 2 factors)
- Mixed designs- at least one factor is between and at least one factor is within (e.g., in pre-post- intervention designs)
- Number of ‘ways’= number of IVs (factors)
- -1-way has 1 IV, 2-way has 2 IV’s etc- only 1 DV in an ANOVA no matter how many ways
Factors/levels
*Factors=IV’s=Treatments
-Levels of a treatment= number of groups/conditions on one factor
-Number of conditions/cells= all combinations of all factors
What is true of the number of factors, levels and conditions in this example? 2 x 4 ANOVA
There are 2 factors, one with 2 levels and one with 4 levels. There are 8 conditions
Where does the variance come from?
WHERE DOES THE VARIANCE COME FROM?
*The effect of a factor= main effect
*The effect of a combination of factors= interaction
(An interaction consists of any variation of the scores which is not due to error or main effects- can be represented graphically where parallel lines=no interaction vice versa)
Interaction plot
Sources of variance in one way ANOVA
Sources of variance in 2-way ANOVA
-Do note taking and review influence each other? e.g., is it going to be more beneficial for people who are taking notes on laptop to review their data, than it is for ppl who are taking notes with pen and paper
*In this design, we have more sources of variance, because each factor adds sources of variance and interactions
Sources of variance in 2-way ANOVA 2
the error is how much people's individual marks vary from the mean or from each other.
The main effect of note taking is how much the mean marks for the three different note taking approaches differ to each other on average.
The main effect of review is how much the mean marks for groups that reviewed vary from the groups that didn't review.
The interaction is looking at how much the mean marks for the different combinations of notetaking and review differ from each other.
2x2x3 design
Sources of variance
*The sources of variance depend on how many factors the ANOVA includes:
-each factor= 1 main effect
-each possible combination of factors=1 interaction
-error is always a source of variance
So how many conditions and how many interactions does a two by five design have?
1 interaction and 10 conditions
How many sources of variance are there in a 2 x 2 x 2 ANOVA?
8
Formula for interactions
- A combination of factors can lead to an enhanced effect or to a reduced effect
- We can use interaction plots (plots of the condition means) to examine interactions
- In interpreting and discussing interactions, we need to explain the observed effects in theoretical terms (e.g., in terms of the scientific mechanisms that can account for the observed pattern in the data).
Describing ANOVA
In ANOVA you might have:
- sig main effects (do they match hypothesis?)
- sig interactions (do they match pattern of data you expected?)
- non-sig main effects
- non-sig interactions
*Every factor, and every interaction, has its own F value, dfs, p, and effect size
*When describing ANOVA in a report, put 2(…) x 3(…) e.g., 2 (type of praise: for effort vs ability) x 3(type of difficulty: for easy vs hard vs impossible)
*When 3 or more conditions in repeated measures IV, you have to check Mauchy’s test of sphericity (look for G-G if this is sig)
example ANOVA output
example ANOVA output 2
example ANOVA output 3
example ANOVA output 4
example ANOVA output 5
example ANOVA output 6
example ANOVA output 7
Qualitative research methods
*The phenomenon or population of the study is the “territory”, and the methods are the “map”
-The methods can tell us something about the phenomenon, but they can't tell us about its completed form
-Different “maps” or methods will provide very different insights
-never get the full picture
-we shouldn't ignore what other methods tell us
*When thinking about qualitative research, we are fundamentally concerned with the moving away from the numerical organisation and interpretation of human behavior to focusing on the words, experiences and sensemaking, researchers are interested in exploring participants talk and very much focusing on the what's, how's and whys
Qualitative research methods 2
*Assumes that meaning is constructed through interactions between human beings within their social world
- no true or valid interpretations but competing ones
- can focus on everyday experiences or unusual crisis
- uses smaller sample sizes but gains richer data and insights
- we want to identify the more and less useful interpretations
History in qualitative research-
Different approaches
CRITICAL PSYCHOLOGY
Fox & Prilleltesky (1997): Psychology is not neutral Research is influenced by power structures Research is a social endeavor and has real consequences Researchers should take responsibility for the moral, social, and political implications of research
FEMINIST PSYCHOLOGY
Interpretation and reflexivity
INTERPRETATION
*To enable researchers to enter the everyday social world to better understand, describe, make sense of, and develop into theories what the socially constructed meanings are that are employed by ppl in day-to-day life
REFLEXIVITY
- Values subjectivity, accepting the role of the researcher on the research process
- reflections are not equitable to biases- you cannot ‘explain away’ your involvement
- there are diff types of reflection:
functional reflection- How your methodological approach to the data may have affected some of the production and analysis of data e.g., data collection and analytic methods, recruitment strategy, your epistemological and ontological position
i.e., some themes may very clearly demonstrate a constructionist or positivist approach to the data that someone else may see differently should they come from a different perspective
More types of relection
Personal reflexivity-
Making yourself visible in the research Positioning yourself in relation to the research i.e., Are you considered to be an outsider to the experiences of the participants, an insider, or something in-between?
You also need to think about how the research has affected you i.e., what have you learnt about yourself during this process?
Critical language awareness-
You want to ensure that your language does not diminish or stigmatize those that you are focusing on
-May not apply to all research, but some subjects in Psychology the use of language and stigma is incredibly important, i.e., ‘fat’ in research looking at obesity
Theory
What do we mean by theory?
*Qualitative researchers are interested in 2 philosophical components- ontology (nature of reality) and epistemology (nature of knowledge)
Theory 2
TA- more researcher led, your theoretical orientations led the analyses
IPA and DA, more theory led, they have more fixed theoretical positions embedded into the approach that the researcher needs to adhere to
-There is a need to describe, explain and apply the theoretical underpinnings of the DA and IPA
Why do we need to think about theory?
*The way we think about reality and knowledge is integral to how we conceptualize and understand the voices of participants
The importance of language
THE IMPORTANCE OF LANGUAGE
Qualitative research is primarily interested in the study of language What is says about the participants What it says about the phenomenon What it says about populations What is says about society and culture
-Therefore, how we, as researchers, communicate and disseminate information is incredibly important -How we discuss our insights and interpretations -How we discuss the research process
Theory 3
Research questions and interview questions
TA- the “what” questions
DA- “how” the knowledge is constructed and represented (deeper level issues, looking at how language is constituting those experiences, and how the nature of language is constructed)
IPA- interested in personal meaning making and sense making (deeper level than TA, how ppants attach personal meaning to a central experience)
preparing for qualitative interviewing:
What is/are your research question(s)? Break this down into ‘mini research questions’ Convert these into ideas for the interviews Research or evidence-led
Reflect: check whether the interview schedule match your research questions Organise the questions in a flow that will enable discussion and allows for flexibility Decide on which questions to ask Cross-check Do not be too rigid
Interview schedules
Constructing interview schedules:
-Begin broadly (Smith et al., 2009) E.g. You’ve agreed to take part in this study because you know someone who has anxiety – Can you talk me through this? -Key Themes -Areas/issues that need further expansion E.g. Is there anything that you wish to talk about that you don’t feel we have covered? -Get their take-home-messages E.g. What do you believe is the most important aspect of your experiences that you think I should know?
Interview schedule or interview guide:
-List of questions to ask in the interview- ensures interview stays on topic and all areas of research interest covered -Semi-structured interviews- flexible use of interview guide -Schedule is For YOU but also for THE PARTICIPANT- write with them in mind (ethically and practically) -Goal is: Exploration - not interrogation!
-Preamble- outlines study, introductory text as a lead into the questions -Questions need to follow a logical order and cluster into topic-based sections -Funnel: general (gentle) to specific (and/or sensitive)
Phrasing interview questions
Phrasing interview questions:
-Ask open ended questions ‘Can you tell me about? ‘How do you feel about?’ ‘What do you think about?’ (Opens up opportunity for them provide a narrative and use their own experience and voice) -Ask non-leading questions- Avoid putting words in their mouths- Instead use hypothetical questions or imagined scenarios e.g., ‘some people think that…’ -Ask clear, short and singular questions- Long and questions that ask multiple things can be confusing- Avoid ambiguity- don’t assume people will understand psychology specific terms -Use prompts- Follow ups to a question e.g., if someone answers yes/no or if you need expansion
-evaluate/review/refine your guide- pilot your project
*Research questions are not interview questions
IPA intro
IPA
-Idiographic approach- focus on the particular (rather than nomothetic approach- claims at group population level)
-Underpinned by critical realist ontology- reality can only be partly accessible and we can come to know some of it through ppant accounts- can get at the conscious processes that the ppant engages in
-an appreciation for the way that culture and language shape the reality experienced by ppants
3 key elements/theoretical origins:
- phenomenology (detailed examination of an individual's lived experience)
- hermeneutics (the analysis/interpretation of experiences’/messages)
- symbolic interactionism (mind and self-emerge from social interactions, and the meaning that these have)
IPA 2
when to use IPA:
-When you’re interested in examining, in-depth, individual and personal lived experiences -Focus on examining how a small number of participants make sense of their personal and social worlds -Works well for social, health, and clinical topics when there’s a need to understand how people make sense of and perceive experiences -When you have assumptions of talk as having a link to their thinking and emotional state, but this is link is imperfect & researcher needs to employ an interpretative lens to delve further (critical realism)
Applying to research questions:
principles of IPA-
- In-depth exploration of personal, lived experiences
- Concerned with questions that hold considerable meaning for the participants
- Can be specific questions or broader
- engage in topics that are current, emotive, and maybe even conflicting
IPA examples
examples-
How does a woman’s sense of identity change during the transition to motherhood? (Smith, 1999)
How do young people with psychosis experience and interpret personal romantic relationships (Shinebourne & Smith, 2009)
What is the experience of deciding whether to take a genetic test or not? (Smith et al., 2013)
DA intro
DA
-A whole approach to psychology and knowledge
-critical of the underpinnings of traditional social psychology that was driven by positivist cognitivist approach
-draws on social constructionism
-DA more concerned with the outside of a person (doesn't assume we can understand internal processes) -simply focused on beh
-instead, language is the level of focus- how they are talking about their experiences
-social norms/social expectations inform the way in which language is used
-explores identity, prejudice, or emotion (which are usually considered in psychology to be internal) as social processes or activities
-purpose and a function for how language is used, and it has social consequences
DA 2
Underpinning theory:
-Social constructionism & Relativism- -DA involves a “strong” social constructionist view of the social world
-Aligned with relativism (multiple and competing realities) Language is constructive and active
-it does things at the social (and individual) level e.g., the law: the law in any one society is constituted by all the statutes that Parliament has passed, all the regulations that are written in the Constitution, and so on *All these are 'just words', but they constitute something very real
-We have access to shared linguistic and cultural resources
-Words have meanings and those meanings construct versions of reality
-We produce and reproduce these pockets of meanings
-We use language as a means to achieve certain ends, or to position ourselves and others E.g., I’m not sexist but….’- to manage others impressions of them
DA 3
when do we use DA:
When you’re interested in looking at how broader social norms and values are produced and reproduced in participant’s talk
-Situate the phenomenon in everyday talk (micro)
-Situate the phenomenon in its broader context (macro)
-Helps to raise awareness to overlooked phenomenon/populations
-Questions tacit knowledge
For example, Health topics, i.e., sexually transmitted diseases Social topics, i.e., poverty and access to healthcare Political and economic topics, i.e., language of nominalization Crime and punishment, i.e., Foucault and the panopticon Clinical settings, i.e., doctor-patient interactions
Applying to research questions
Applying to research questions:
principles-
- Research questions focus on construction, rhetoric, ideology, and action
- movement to the function and action of language
How are ME/CFS (and other labels) constructed as illness categories by GPs? (Horton-Salway, 2007) How is evidence of serious adverse effects from SSRIs is managed by those who have a professional stake in using or promoting the drugs? (Libert & Gavey, 2009) How do young adults talk about and produce meanings and understandings of immigration, immigrants and cultural diversity? (Lyons et al., 2011)
Ethics
ETHICS
-Firm but flexible application of ethics: core principles not a barrier, a safety net and a discussion
-aim is to produce quality research
Marginalized voices
Marginalized voices 2
Voice and representation
Voice and representation 2
Voice and representation 3
Ethics in the room
Ethics in the room 2
Ethics and honesty
Overview
Qualitative data collection
what is qualitative data?
-Qualitative data is where we use words as data Focus on the meaning and impact of those words
-Most common form of qualitative data is talk Usually generated through purposeful interactions with participants
-Other types of data can include: Self-written documents Historical documents Public-facing documents Advertising Speeches Videos Images…
Qual methods
The key thing is that they are all interested in language and how people or persons communicate their experiences, perceptions and understandings.
Qual methods 2
DECIDING WHICH METHOD TO USE
-No ‘right’ or ‘wrong’ – focus is on selecting the most appropriate method -This should be a considered to be a holistic process -Focus should be on selecting a method that helps answer the research question. -There are some methods that afford flexibility and others which do not
*Braun and Clarke highlighted some questions to help guide this decision:
Focus groups
FOCUS GROUPS
*Directed conversation with more than one participant
-Aim to find out as much as possible about the participants’ understandings and meanings, with more than one participant! -Individuals come together to discuss a topic Involves sharing of experiences, ideas, views etc.
-Between 3-6 participants (Barbour & Kitzinger, 1999)
-size and number of focus groups impacted by number of factors including homogeneity of participants, focus of topic
Why use focus groups?
- Contextualizes collective understandings and sense-making (Barbour, 2007)
- Useful in considering peoples’ shared understandings (Kitzinger & Barbour, 1999)
- Sensitive to points of consensus and disparity
Setting up focus groups
Focus groups 2
What happens?
-Work through the interview schedule/activities
-Ask for others’ opinions- Ask for examples
-Allow participants to question each other, respectfully
-Be willing to follow the conversation
-Keep an eye on the time and wait for a lull in the conversation
-Ask if they have anything to add- ask if they have any questions for you
-Thank them for their participation
-Ensure they take the information/debrief away with them
Focus groups 3
Troubleshooting
- Thoughts on contribution-does everyone need to contribute, and do so equally? -Realist/essentialist- there would be a hope of equitable contribution from all ppants -Relativist/constructionist-they are not simply vessels of information- way ppants behave is part of a constructionist conversation- equal contribution is not necessary
- The expert- may either dominate or tell the research what the facts of the project should be. -Resolved by reminding participants about the importance of their insights
- Chatterbox- somebody who dominates the focus of conversation -Resolved by: Use non-verbal cues, i.e., decrease eye contact, turning body away, -Resolved by asking others what their views are
- Shy/quiet participant Difficulty: You don’t want to overstep the boundaries and ‘force’ responses -Resolution: focus on directly involving them more (but try to avoid over-doing this), i.e. Do you have anything you'd like to share? -Resolution: use non-verbal cues, i.e., making eye contact, smiling, turning body towards them
- Bored participant -Resolution: try and engage and use similar strategies as you would a shy person
online focus groups
ONLINE FOCUS GROUPS
*The format may be different, but the content seems relatively stable between F2F and online Focus Groups (Woodyatt et al., 2016)
Asynchronous Online Focus Groups
- more time to think about responses- create a more empowered experience
- there could be technological issues associated with them
Synchronous Focus Groups- real time- like face to face
- technology can provide different types of environments for participants to engage with
- requires a good, and consistent bandwidth, and reliant on individual schedules
Groups in the “virtual world”- simulated environment
- avatars may lead to greater engagement and co-creation activities
- assumes a certain level of skill/ability is needed to engage in environment successfully
Benefits of focus groups
Benefits of focus groups
- Useful in considering peoples’ shared understandings (Braun & Clarke, 2013)
- Sensitive to points of consensus and disparity (Kitzinger, 1994, 1995; Wilkinson, 1998)
- Socially-situates social phenomenon; “people do not operate in a social vacuum” (Kitzinger, 1994, p. 112)
- Flexible to unexpected topics (Braun & Clarke, 2013)
- Participants may be more willing to speak alongside others similar to them (Liamputtong, 2007)
- Decreases the power of the researcher (Braun & Clarke, 2013)
Drawbacks of focus groups
- Recruitment can be difficult
- Arranging times can be difficult
- Hard to manage interaction
- Transcription can be more complicated and time consuming (Braun & Clarke, 2013)
- Group must share some similarity- may make it difficult to recruit ppants if you have a very specific focus (Acocella, 2012)
- Difficult to explore individual experiences/sense-making (Braun & Clarke, 2013)
Ethics and sensitive topics
Ethics
Key risk: confidentiality -Whilst there is a risk to participants discussing their experiences of the focus group (Wilkinson, 1998)
- Addressing this risk:
- Create a series of ground rules (Farquar & Das, 1999) to ensure respectful discussion in the focus group, and that no information about other participants is revealed which may make others identifiable
- Ensure a distress protocol ‘works for focus groups, i.e., is it as easy to leave a room?
Sensitive topics
*Focus groups might appear to be a counterintuitive method for encouraging discussion about sensitive topics due to the presence of others
- However, If conducted with respect and reverence for the other participants, focus groups can be a lucrative form of data collection (Braun & Clarke, 2013; Wellings, Branigan, & Mitchell, 2000)
- Focus groups can increase self-disclosure, rather than inhibit it (Farquar & Das, 1999)
- Potentially even disclosing potentially “discrediting experience(s)” (Wilkinson, 1998, p. 192)
- Participants will outnumber researchers in focus groups, so may be more empowering than interviews (Farquar & Das, 1999)
- The presence of others may encourage quieter participants to come forward (Kitzinger, 1995)
focus groups summary
Participation and power
*Power doesn’t just exist within or because of the research setting- it exists within society more broadly
-presents itself in research
Ethnography
ETHNOGRAPHY
- Tradition of ‘naturalism’-mundane everyday experience
- Form of observation
- Usually resulting in large amounts of qualitative data
- Extended periods of data collection through variety of formats (surveys, focus groups, observations, and field notes)- time intensive and resource intensive
- May require the researcher to exist within the social world of the participants- can be covert or overt
- Want to appreciate the phenomenon in its own social world
- Underpinned by reflexivity and deep awareness what researcher is bringing to research
Ethnography 2
When to use ethnography- 3 key times:
-When you want to understand the way that naturally occurring group's function Reliant on the skills of the researcher if overt
-When broad observations are of interest i.e., not wanting to address topics/foci that would be better suited to fine-grain and individually focused meaning making like IPA
-Way to explore the culture of participants (within the context of the study, i.e., cannot get ‘full picture’ of their whole life)
Ethnography 3
Applications of ethnography:
-Healthcare e.g., how do nurses respond to crises on a maternity ward, for example,
-SCB- context in which ppl consume these substances
-C/D groups- groups that would be quite wary of external party’s
-organizations- e.g., on bouncers- it gives you sort of like a broader insight into sort of group dynamics, culture, everyday activities of the people who are involved in the in these.
issues in ethnography
- Issues in ethnography
- The focus is on micro-interactions: what about macro concerns? - broader social environment
- Whose interpretation is being prioritised?- often the researchers voice which comes through the strongest but should be heavily data driven
- How do you access your participants? Are there gatekeepers? - could contaminate the research
- Resource-intensive
- Ethics
- Harm? Would there be harm resulting from the project? What if people are identifiable unintentionally? If covert, what harm would be done?
- Informed consent? If covert, then consent might be difficult/impossible to gain without influencing the research Invasion of privacy? If covert, would it be considered to be an invasion of privacy, that the things that you’re saying to someone is going to be used in research? How would you feel if someone recorded your chat over a coffee and wrote about it/inferred things about you from it?
- Deception? If covert, are they being deceived if they believe you to be a member of their community-influences types and nature of discussions and disclosure Do the end justify the means? More of an issue with covert research, overt ethnography would mean participants would be informed and consented.
Action research
ACTION RESEARCH
Helps to empower and out power back into hands of those being researched
(Qualitative) research has been criticized for marginalizing the voices of those who the research aims to ‘empower’
-Critical Qualitative researchers need to be aware of issues of power, who has power, who is empowered, and who are disempowered
Action research:
-Action research troubles the notion of power and researchers as enactors of change- centralizes the populations of focus
-Systematic approach to investigation that enables people to find effective solutions to problems they confront in their everyday lives (Stringer, 2007)
-Action research aims to work with populations- identifying issues and bringing about resolutions and in that process, that population has become empowered,
Principles of action research
*Very different to traditional qualitative research where participants are only involved in process when you are actually collecting the data
LEWINS MODEL OF ACTION RESEARCH- A cycle and spiraling action revision effect
*Iterative process- constantly looking back and adjusting plan accordingly
Theory of PAR
THEORY OF PAR
Informed by the theory of Paulo Freire, The Pedagogy of the Oppressed (working with people who sit in oppressed social or environmental positions)
-‘banking’ concept of education (where students seem to sort of passively accept knowledge in society and those in power dictate what society should look like)
-Conscientization (acknowledging the social reality that's imposed on people)- showing that ppl can change this
*The process encourages critical thinking by allowing the community to take an active role in the process of generating knowledge
Applications of PAR
Methods and applications:
*All used as tools for empowerment- ways of getting insights and feedback and thoughts to influence the change, the action that's hoping to come from the research itself.
Diagramming Mapping Community art and media Surveys Focus groups Interviews Diaries Photovoice World café
Photovoice and world cafe
- Photovoice
- Photography to explore people’s worlds and make them accessible to others- ppants encouraged to take photos then these images are discussed in interviews
- Encourage documentation and reflection -Empowerment though personal and shared experiences -To encourage critical dialogue -To speak to those in powerful positions, i.e., policy makers
- World cafe
- Multiple themed tables in a larger setting Usually, circular tables with chairs-Evokes the café feel!
- Participants discuss the topic -Sometimes have prompts/facilitator
- Move around the room, talking about different themes
- Recorded via Dictaphones, but people allowed to draw/write on tables/post-it notes
- Networks of conversation
- Focus on constructive discussions and building relationships -Focus on ownership of views and bringing up new insights
- Emphasis on ‘local knowledge’ (p. 212)
- Cross-pollinate ideas across tables- the knowledge gained through discussions deepens and the ideas and the learnings from previous discussions is brought to bear in the discussions on subsequent tables and subsequent topics.
PAR and ethics
Interviews
INTERVIEWS
*Aim to find out as much as possible about the participants’ experiences and meanings
*semi-structured interviews are the only ones really used- interview guide has been informed by the literature
Qualities of interviewing
QUALITIES OF INTERVIEWING
Gathering rich data Detailed, focussed, full Help construct what lies beneath the surface: participants’ experiences, views, perceptions, and contexts How? Non-leading questions Attitude of ignorance and naivety (Willig, 2001)
Establishing rapport A relation of mutual understanding and trust How? No recipe! Attitude of respect and curiosity Not correcting the participants’ accounts Comes with practice!
*In procedure, you need to make sure it is replicable- e.g., were you comfortable with silence, what was your demeanor etc.
Language and meaning Meaning through what is said Meaning through pauses, context, silences, non-verbal expressed emotions How? Careful ‘listening’
Diversity Differences between interviewer and interviewee Race, gender, class, race, age, language, religion, sexuality How? Reflexive approach
Interviews 2
Online interviews
ONLINE INTERVIEWS
What research questions can I use? Similar to f2f though, there may be more scope to explore controversial/sensitive topics which may be awkward to discuss face to face
Recruitment This can be a little broader, and may attract those normally put off by research You may be able to gain access to a more diverse sample
Process Very similar to f2f interviewing Participants have greater control
Flexible scheduling Allows for the visuals Which could improve rapport (Mirick, & Wladkowski, 2019) Understand non-verbal communication (Knapp, Hall, & Hogan, 2013) Only-audio may be preferable, esp. around anonymity (AlKhateeb, 2018)
Ease of capture Interview spaces can be a lot more flexible, May provide a more empowered experience for the participants (Mirick, & Wladkowski, 2019) Participants have greater control
Tricky customers
TRICKY CUSTOMERS
- Silence Probe for further examples, expansion, and clarification- Draw on more of the closing questions-these open allow for scope from the participant
- Monosyllabic responses Ask for expansion, examples, clarification Try to rephrase the questions (if they’re unsure)
- Chatterbox Both on-topic and off-topic Allow them to finish what they are saying, thank them for their contribution, & try to steer them back on-topic
- Heightened emotions Manage self-disclosure Allow them to finish what they are saying & try to steer them back on-topic Remember, if you are uncomfortable, the interview can end, politely
Confidentiality
Qualitative analysis
- A way of creating meaningful patterns in the data- There is a spectrum of analytic approaches
- Qualitative analyses are systematic ways of generating meaningful insights
- Approaches tend to prioritise certain elements of participants talk
Qualitative analysis 2
content- superficial content of data- ****Sometimes it's also uses coding frameworks to count the frequency of a given predefined behavior.
Thematic- common, focuses on identification of themes on a range of topics
Discourse- considers talk as a function of social action, and that language is used purposefully- discourses ****represent patterns in the data that reflect something of the constricted nature of the phenomenon.
Conversational- focuses on how interactions are represented by talk- and what action that talk represents in naturally occurring conversations, i.e. the process of interpretation, how it's managed and constructed.
IPA- methodological approach in itself, has series of philosophical assumptions that underpin it- focuses on individual sense-making, experimental knowledge and attempts to understand participants experiences from their perspective
Grounded theory- interested in the identification or confirmation of theoretical models of human ideas
Analytic techniques
- Different analytic techniques-
- prioritize different aspects/elements of participant speech -are better suited to answer different research questions -work better for different types of data
- Some techniques are more flexible to researcher positionality, others come with more ‘fixed’ elements
Which analysis?
No ‘right’ or ‘wrong’ – focus is on selecting the most appropriate method This should be a considered to be a holistic process Focus should be on selecting a method that helps answer the research question. There are some analytic techniques that afford flexibility (TA) and others which do not (IPA)
Preparing for analysis
Make sure your data is stored adequately, Make sure your data is prepared for analysis (transcribed), Have your epistemological and ontological position clear, Make sure you understand the process of analysis you are undertaking, Ensure that this type of analysis can answer your research method and is suitable to your data, Decide how you might want to conduct your analysis and you have the right tools, Give yourself time!
IPA analysis
IPA
Qualitative Methodology in its own right Focus on personal sense-making of lived experiences This sense making is recognized as an interpretative enterprise Useful for examining topics which are complex, ambiguous, and emotionally laden
E.g., life transitions or psychological event
-suited to clinical and health settings
PHENOMENOLOGY
The study of phenomenon, specifically the structures of experiences How participants make sense of their lived experiences while embedded in their personal and social worlds Interested in understanding personal perceptions and accounts of the experience Not attempting to produce an objective statement of the object or event
IPA analysis 2
HERMENEUTICS
The theory of interpretation People are interpreting (making sense) of their experience (links to critical realism) The researcher is then interpreting their interpretation
Double hermeneutic “The participants are trying to make sense of their world; the researcher is trying to make sense of the participants trying to make sense of their world” (Smith & Osborn, 2003, p. 15) Highlights the central role of the researcher in the research process You can also do this curiously/critically, i.e., ‘What is the person trying to achieve here? ‘Is something leaking out that wasn’t intended?’, but always empathetically Assumption that participants have internal processes, but may not always be able to communicate this well-role of the researcher to interpret and ‘fill in the gaps’ of internal processes from their talk
IPA critical realism
CRITICAL REALISM
IPA conceptualizes the person as having internal processes (i.e., cognition, affect) and there is a connection between their talk and internal state
However *This connection is not always clear and direct; people may struggle to articulate their thoughts and feelings & the researcher is responsible for helping bring clarity to those processes
We cannot ‘get at’ the ‘real’ experience/event, instead we understand the experience/event through the participant’s account, Accounts represent something of the reality of that experience/event However, the accounts are influenced by the personal and social lenses of the participants and are limited by their linguistic practices Thus, the accounts can tell us something of the nature of the phenomenon, but not its true form
Further influencing principles
CONDUCTING AN IPA
Further basic principles of IPA: Assumes agency of the individual (have control over way they phrase things) Primary focus on understanding individuals’ lived experiences and how they make sense of those experiences Dynamic interpretive endeavor
Research Question Tend to be open-ended with the purpose of gaining a rich description of the phenomenon Focus on a significant event e.g. How do people make sense of the experience of being a single father?
planning IPA
Planning:
Sampling Tendency for smaller sample sizes Focus on homogeneity A number of factors to be considered Focus on detailed accounts of experience
Selecting the method Need to gain detailed accounts Usually semi-structured interviews Interviews provide the space to explore sense-making practices Focus Groups? Observations? likely to be difficult at accessing individual languages
Stages of IPA
*At stage 1, you are trying to understand the main elements of participant accounts
*Emotional reactions can be neg and pos
Stages of IPA 2
Stages of IPA 3
Stages of IPA 4
Stages of IPA 5
Stages of IPA 6
IPA overview
Discourse analysis
*Discourse is a focus on language and meaning *Communication by words, talk and speech; a linguistic unit longer than a sentence (unlike TA)
*Discourses as linguistic constructs/texts which shape how we can think and talk about a given phenomenon, conceptual understanding of reality *DA involves a “strong” social constructionist view of the social world *DA is not simply a set of techniques for conducting research; it also involves a set of assumptions concerning the constructive effects of language. *Reflexivity and the role of the researcher is central.
What are discourses?
Language and social constructionism
Discursive practices
Language as functional
Linguistic features
Conducting DA
Conducting DA 2
Conducting DA 2
Conducting DA 3
Conducting DA 4
Conducting DA 5
Comparing IPA and DA
Reflexivity in IPA
Reflexivity in IPA 2
DA and data
Why use discourse analysis?
Why? Situates the phenomenon in everyday talk (micro) Situates the phenomenon in its broader context (macro) Helps to raise awareness to overlooked phenomenon/populations Questions tacit knowledge
Where? Health topics, i.e. sexually transmitted diseases Social topics, i.e. poverty and access to healthcare Political and economic topics, i.e. language of nominalisation Crime and punishment, i.e. Foucault and the panopticon Clinical settings, i.e. doctor-patient interactions
Other directions in DA
Discourse Analysis and alternative data
Public documents, speeches, policies Diaries, blogs, books Media and internet Discussion boards Social networking YouTube Posters, leaflets, and other campaign materials
Other types of DA
Other types of DA 2
DA evaluation
Quality and critiques
Quality in qual research
Addressing quality
Addressing Quality(Creswell & Miller, 2001)
Triangulation: Methodological triangulation Use multiple methods (collection & analysis) on the same population/phenomenon Consistency in themes = greater trustworthiness of findings
Researcher triangulation Multiple researchers code the same data Consistency in themes = greater trustworthiness of findings
Audit Trail: Field notes- Taking notes after the interviews/focus groups Identifying issues, you felt were particularly pertinent to the participants
Clear and evidenced analytic process Retain & outline analytical process In whatever shape that might be
Audit Trail- Clear and justified procedure Make every decision clear Justify those decisions with literature
Deviant cases Participants accounts that ‘don’t fit’ Their voices still need to be recognized
Justification
Reflexive practice
Reflexive Practice Clear and considered design and dissemination Underpinning consequences Analysis of data Well-supported conclusions
Reflexivity Process throughout the research process What you have brought to the research Thinking about how the research may have affected and possibly changed us, as people and as researchers
Validity procedures
*Types of quality assessments- diff types that work with diff types of approaches in qualitative research
Common qual critiques
*don’t use these!
The spectre of objectivity
The spectre of objectivity(Morgan & Drury, 2003)
Qualitative research is critiqued for lacking objectivity Objectivity is associated with limited researcher-influence Objectivity is considered to be a marker of ‘good’ scientific enquiry
Social constructionism in action… Objectivity has come to mean truth, fact, neutrality, and reality Subjectivity has come to mean tentative, less-than-real, biased, non-useable
In qualitative research… Subjectivity is embraced Subjectivity is acknowledging and exploring the interwoven role of the researcher, the studied, and the socio-historical-political context in the research process Subjectivity is an accepted aspect of qualitative research It is not seen as a limitation, It is an acceptable and essential aspect of qualitative research Addressed through methodology & reflexivity
Generalizability
Generalizability?
Can qualitative research not be generalised?
While generalisability is always a goal, real-world research is often less generalisable than we would like (even quantitative research). A single study cannot be easily generalised, rather evidence is cumulative & produces a body of evidence. Researchers talk about arriving at a scientific consensus within the bounds of the empirical evidence. The process of research (and science) is also one of continual competition and criticism—the continuous testing of the consensus#
Sample sizes in qualitative research tend to be smaller than in quantitative research Sample size is a key criticism levied at qualitative research This is particularly relevant to the notion of generalisability and reliability There needs to be an adequate and detailed justification for sample sizes
Ceasing data collection
Ceasing data collection
Getting the balance right between enough data vs unwieldy data! (Sandelowski, 1995)
Saturation focus on no new insights being developed. Vague and impractical (O’Reiley & Parker, 2012) Not consistent with critical qualitative approaches (Braun & Clarke, 2019) Fails to address issues like diversity, theoretical context, and specific qualities of the study (Sim, Saunders, Waterfield, & Kingstone, 2018)
A priori sample size decisions -Rules of thumb -Numerical guidelines -Statistical formulae
Issues with these approaches: -Ontological status of the theme Themes as ‘instances’ Analytic context Diversity of participants Interdependent determinants Statistical assumptions Assumption of generality
Ceasing data collection 2
Methodolatry
Mixed/multi-method design
Mixed/Multi-method design
Qualitative methods used alongside/to support quantitative methods (Lambert & Loiselle, 2008)
Mixed-methods approach, with one key study with other supplementary studies (Morse, 2012)
Using multiple methods for means of triangulation (e.g. Frost et al., 2011)
Multiple methods of data collection and analysis to form one singular results section (e.g. Clarke, et al., 2015)
Issues relevant to using multiple qualitative methods, i.e., epistemological and ontological consistency (e.g., Chamberlain et al., 2011)
Comments
No comments have yet been made