PS20074: RESEARCH METHODS AND RESEARCH DESIGN: RESEARCH TECHNIQUES
INTRODUCTIONDr Chris Griffin
Basic terms in research
TermMeaning
ModelAn overall framework for looking at the world
ConceptAn idea deriving from a given model
TheoryA set of concepts used to define and/or explain some phenomenon
HypothesisA testable proposition
MethodologyA general approach to studying research topics
MethodA specific research technique (eg. interviewing, questionnaires)
Common assumptions (‘traps’) in social research
- Scientism = the uncritical assumption that ‘science’ is superior to ‘common sense’.
- Progress = 19th C. western belief in the inevitable march of progress in history, aided by scientific discovery.
(1) and (2) mainly found in quantitative research
- Tourism = when research are akin to upmarket tourists, in search of excitement and novel or spectacular experiences.
- Romanticism = 19th C. western belief that the arts and music express the true inner world of the artist and the audience. In research, belief that interviews etc. will be able to reflect the authentic inner worlds of participants.
(3) and (4) mainly found in qualitative research
Four Main Methods in Social Research
MethodQuantitativeQualitative
Observ-PreliminaryFor understanding
ationworkanother culture
TextualContent Understanding
analysisanalysisP’s categories
InterviewsSurvey Open-ended
researchQ’s to small
(fixed-choice Q’s)samples
Audio andUsedUnderstanding
videoinfrequentlyhow P’s organise
recordingtheir talk and
body movements
Main Methods in Quantitative Research
(Bryman, 1988)
MethodFeaturesAdvantages
SocialRandomRepresentative
surveysamples,+
Measuredtests hypotheses
variables
ExperimentExpmtalPrecise
stimulus,measurement
Control group not
exposed to stimulus
OfficialAnalysis of Large
statisticspreviouslydatasets
collected data
StructuredObs. recordedReliability
observationon preplannedof
scheduleobservat’s
ContentPreplanned Reliability
analysiscategories usedof measures
Limitations of Quantitative Research
- Can involve a ‘quick fix’, little contact with people or the ‘field’
- Statistical analyses may be based on variables that are arbitrarily defined
- After the fact speculation about the meaning of statistical analyses can involve the sort of ‘common sense’ reasoning that science tries to avoid
- The pursuit of ‘measurable’ phenomena can mean that unnoticed values enter the research, by employing problematic concepts such as ‘delinquency’ or ‘intelligence’ in an uncritical way
- A purely statistical approach to research can result in the generation of further, increasingly trivial hypotheses from existing data
The Preferences of Qualitative Researchers
(from Hammersley, 1992)
- A preference for qualitative data: understood as the analysis of words and images rather than numbers
- A preference for naturally occurring data: for observation rather than experiment, unstructured or semi-structured rather than structured interviews
- A preference for meanings rather than behaviour: attempting to document the world from the perspectives of the people studied
- A rejection of natural science as a model
- A preference for inductive, hypothesis-generating research rather than hypothesis-testing
Four Approaches to Qualitative Research
ApproachConceptsPreferred data
NaturalismActorsObservation
MeaningsInterviews
Ethno-P’s methodsAudio /
methodologyfor assemblingvideo
phenomenarecording
EmotionalismSubjectivityInterviews
Emotion
Post-RepresentationAnything
modernismReflexivitygoes
PS20074: RESEARCH METHODS AND RESEARCH DESIGN: RESEARCH TECHNIQUES
INTERVIEWINGDr Chris Griffin
Approaches to Interview Data (Silverman)
ApproachStatus of dataMethodology
Positivism‘Facts’ aboutRandom
behaviour &samples;
attitudesStandard
questions;
Tabulations
Emotionalism‘Authentic’Unstructured,
/Humanismexperiencesopen-ended
/Naturalisminterviews
ConstructionismMutuallyAny interview
constructedtreated as a
topic
Positivism and Types of Knowledge
1)‘Facts’: eg. biographical information, statements from informed sources that can be checked for ‘accuracy’
2)Beliefs about ‘facts’: these may vary between individuals
3) Feelings and motives: many positivist researchers advocate the use of open-ended questions using respondents’ own language
4)Standards of action: ie. What people think should or could be done in a given situation. May involve questions on hypothetical situations if respondent has not experienced a given event
5)Present of past behaviour: as in (4) responses are assumed to be more accurate if respondent has experienced the situation themselves
6) Conscious reasons: beliefs respondents have about causes of a particular phenomenon.
Positivism: The Truth is out there
**Positivists aim to elicit a body of ‘facts’ or information that is assumed to already exist ‘out there’ in the world
**They aim to make their interview techniques and research designs as reliable as possible in order to avoid the possibility that interview responses might be shaped, in part, by the interview setting itself.
**This leads to an emphasis on standardised interview schedules, highly structured interviews and emotionally distant interview technique
**Positivist researchers aim to be objective, apolitical and value-free.
Emotionalism/Humanism/Naturalism
**Emotionalism aims to elicit accounts of subjective experience that are ‘authentic’ for that respondent.
**They favour ‘unstructured’ or open-ended interviews, which aim to allow respondents to communicate their authentic experiences undistorted by the views or interference of the researcher.
**The aim is to develop a close rapport between respondents and interviewers, with interviewers sometimes telling their own stories to respondents. The aim is to break down (to some extent) the hierarchical relationship between researcher and researched (eg. in some feminist research)
**Unlike positivism, emotionalism acknowledges (sometimes encourages) emotional ‘empathic’ connection between respondents and interviewers.
Constructionism
**Constructionists view respondents’ accounts as “not simply representations of the world [but] part of the world they describe” (Hammersley and Atkinson, 1983)
**Respondents are assumed to actively create knowledge in collaboration with the researcher in the context of the research interview. The emphasis is as much on how this knowledge is assembled (constructed) as on what is said in the interview
** Constructionists aim to show how interview responses are produced in the context of the research interaction between researcher and respondent, without losing sight of the meanings produced in the interview encounter or the circumstances that shape the meaning-making process.
** Their analytic procedure aims to describe how talk/accounts are produced in specific (interview) situations, but also to show how what is being said relates to the experiences and lives of the respondents
Approaches to Interviews
a)Interview as Technique
(Positivist approach)
**Standardised sets of questions are seen as part of good interview practice, designed to ensure that the interview is a reliable research instrument, free from interviewer ‘error’ or ‘bias’
**Any similarities in respondents’ answers are explained in terms of ‘face-sheet variables’ such as social class, gender or ethnicity, that are seen as external to the interview context
b)Interview as local accomplishment
**Standardised questions derive their sense from commonly available conceptions about people’s behaviour and the social world. It is not possible to ask a truly ‘neutral’ question
**Similar answers relate to the respondents’ skills in deploying shared knowledge about this shared social world.
Interviewing Skills
1)Asking
2)Listening
3)Interpretation
All of which require open-mindedness and concentration.
None of these skills come ‘naturally’, they all require training and practice
All three are theoretical projects that are influenced by the theoretical approach taken by the researcher in any given project.
Key points to remember
**Your choice of research method should always suit your research question(s) – and so should your choice of analytic procedure
**Although interviews do not enable researchers to gain access to ‘naturally occurring data’ about the world, this does not invalidate interviews as a research technique. Everything depends on the status we accord to the data/information gathered in interviews, and how we interpret it.
**It is never possible to discover whether interview responses give researchers access to ‘true’ or false’ reports about reality. It is more useful to treat such responses as displays of perspectives or moral positions by respondents in specific interview contexts.
Interviewing
Possible reasons for using interview techniques:
* Concern with subjective meanings of respondents
* Allows exploration of complex and contradictory issues
* Provides a lesson in research involvement and practice
* Can examine the operation of power relations in research
(Banister et al, 1994)
Useful Questions to ask about your study
1) Why (and in what way) is your chosen topic significant?
2) How far do your topic and findings relate to other research?
3)Why is an interview method appropriate for your topic and your research question?
4)Is the size and method of recruitment of your sample appropriate to your topic and your model?
5)Will you audio or video record your interviews? How/will you transcribe them?
6)Will you interview your respondents face-to-face? (or use email?)
7)What status will you accord to your data? Are you seeking objective ‘facts’, subjective ‘perceptions’ or ‘narratives’?
8)How thoroughly will you analyse your data? Will you report a few ‘telling’ or ‘typical’ extracts? How will you decide which extracts to report? Have you worked through all your interview material, searching for examples that don’t fit your original suppositions (‘deviant case analysis’)?
Interviewing in Practice
a) Select research topic and rationale (ie. why do this study), including your research question(s)
b) Select respondents (ie. decide who to interview) in terms of which groups or individuals best exemplifies the range of perspectives relevant to your research topic and your research question(s)
c) Generate your interview schedule and pilot this with a similar set of respondents (ie. decide what to ask)
d) Contact prospective respondents; consider how you present yourself and your research study; ask their permission to take part in the study
d) Negotiate a research contract, involving one or more of the following:
** a guarantee of anonymity or confidentiality
** promise to stop at any time on request
** exclude from the data anything the respondent wishes
** provide copy of research results/ report
** input into research analysis and /or written report by respondents
Qualitative Data Analysis:
1)First Stage
a) Code and mark your data into overall categories or themes
b) Code and mark your data into sub-categories under each overall category heading. Some sections of the data may fall under more than one category.
c) Begin to interpret your findings in relation to your research question(s)
2)Second Stage
a) Re-order your categories in relation to the argument you want to present in your analysis; this can involve ‘cutting up’ and re-ordering your data
b) Further interpretation of your data in the light of your research question(s)
c) Find a study using qualitative research methods that employs a style that you feel is appropriate for your data. Use this to provide guidelines for your written report
d) Write up your data in a similar format to the ‘model’ study above - and be prepared to revise it if necessary.
NARRATIVE ANALYSISDr Chris Griffin
(From Lawler chapter in May: ‘Qualitative Research in Action’)
Narrative analysis:
- focuses on “the ways in which people make and use stories to interpret the world”
- does NOT treat narratives as stories that transmit a set of facts about the world, and is not primarily interested in whether stories are ‘true’ or not (so is closer to social contructionism than positivist approach)
- views narratives as social products that are produced by people in the context of specific social, historical and cultural locations
- views narratives as interpretive devices through which people represent themselves and their worlds to themselves and to others
Narrative theory argues that:
- people produce accounts of themselves that are ‘storied’ (ie. that are in the form of stories/narratives)
- the social world is itself ‘storied’ (ie. ‘piblic’ stories circulate in popular culture, providing means people can use to construct personal identities and personal narratives). Ricoeur argues that narrative is a key means through which people produced an identity.
- Some of most interview accounts are likely to be ‘storied’ (ie. in narrative form)
- Narratives link the past to the present, but …
- There is no ‘unbiased account of the past
Definitions
Narrative can be characterised by:
- Accounts which contain an element of transformation (ie. change over time)
- Accounts containing some kind of action and characters
- That are brought together in a plot line
So:
- narratives have a temporal dimension
- characters and actions can be imaginary/fantasy
- ‘emplotment’ is a process through which narratives are produced: many disparate elements go together to make up one story (eg. digressions, sub-plots etc.)
- Narratives must have a point (a ‘so what?’ factor), which often takes the form of a moral message
Research Methods and Narrative Analysis
Research that focuses on the role of narrative:
- Usually involves life story research or oral history
- Usually adopts a qualitative approach, using semi-structured interviews rather than questionnaires
- Usually the researcher says very little, acting primarily as an attentive listener, but …
- All narratives are always co-constructed, even if the audience is oneself or an imaginary other, or if the story is told to oneself in the form of a daydream
Structuralist approaches to narrative:
eg. Propp, 1968 / Labov, 1973
(from Silverman’ 2nd edition, ‘Interpreting Qualitative Data’)
Narratives can take different forms, and Propp (1968) argued that:
- The Fairytale involves a narrative form that is central to all story-telling
- The Fairytale is structured not by the nature of the characters but by the function they play in the plot
- And the number of possible functions is fairly small
Example:(Using Propp’s approach)
Most fairytales follow a similar plot line…
‘A dragon kidnaps the king’s daughter’
ElementFunctionReplacement
DragonEvil forceWitch
KingRulerChief
DaughterLoved oneWife
KidnapDisappearanceVanish
Now – can you do the same using ‘Star Wars’ as an example?
Narrative Theory: Approaches to the study of narrative
(a partial and incomplete list)
a)Structural analysis: eg. Labov, 1973
Focus on story grammar
b)Sociology of stories approach: eg. Plummer, 1996
Focus on cultural, historical and political context in which particular stories are (or can be) told by whom and to whom (eg. ‘coming out stories’)
c)Functional approach: eg. Bruner, 1990
Focus on what work particular stories do in people’s lives
------
Examples of structural analyses of narrative
Labov, 1973Stein, 1979
1)Setting/ orientation1)Setting
Abstract/ summary of story
2)Initiating event2)Initiating event
3)Complicating action3)Internal reaction/
response of protagonist
4)Resolution/ result of action4)Action by
protagonist to deal
with situation
5)Evaluation/ point of story5)Consequence of
action
6)Coda/ return speaker to 6)Reaction to events/
presentmoral of tale
Bruner, 1990: ‘Acts of Meaning’
**Functional analysis of story-telling as a means of conveying meaning
**Functions of narrative = solving problems
= tension reduction
= resolution of dilemmas
**Narratives allow us to deal with and explain mismatches between the exceptional and the ordinary. When events occur that we perceive as ordinary, then explanations are not required.
**Narratives allow us to re-cast chaotic experiences into causal stories in order to make sense of them, and to render them safe.
------
Canonical Narratives
**Narratives of ‘folk psychology’ (or ‘common sense’) summarise ‘how things are’ and (often implicitly) how they should be.
**When we perceive that things are ‘as they should be’, the narratives of folk psychology are unnecessary.
**Narratives are a unique way of managing departures from the canonical
------
Practical Exercise / Interview Assessment
For the interview assessment, students are required to conduct, transcribe and analyse an interview, using one of the data analysis procedures covered in the unit. Students will be required to present their research question(s), interview schedule, and demonstrate that they understand how the interview schedule will answer their research question(s). They will also present a brief basic analysis of the interview aterial, discussing this in relation to their research question(s) and commenting on the strengths and weaknesses of their interview study. [25% of total assessment of the unit].
The deadline for handing in this piece of work is Friday January 16th 2004
TOPIC for Interview Assessment:
Students’ accounts of the meanings associated with teenage drinking (or with drinking now)
** This practical involves the use of semi-structured interviews with individuals about the meanings associated with teenage drinking / drinking now.
** You should approach friends, house-mates and/or fellow students to act as respondents in this study. Please do not interview other Psychology students as part of this practical exercise.
** Respondents will be giving retrospective accounts of their use of alcohol (now or during their teenage years), so the data you collect should be treated with some caution. These interviews will not provide you with accurate information on respondents’ drinking habits as teenagers, nor about possible influences on their alcohol use.
** Qualitative methods, and especially semi-structured interviews, are particularly valuable as a means of finding out about how respondents construct the meanings associated with drinking (ie. what they think drinking meant to them as a teenager or what it means to them now), but from the perspective of their age at the time of interview.
Stages involved in the Interview Study:
1)Decide on the precise aspect of the topic you want to study. Students can work on their own, in pairs or in small groups (no more than 4), but each student must conduct one interview, and analyse and write up the interview material themselves.
2)Decide on your research question and who to interview (eg. males and females, or only females, only males etc.)
3)Decide on the form of analysis you wish to employ (eg. Grounded Theory, Discourse Analysis, Narrative Analaysis)
4) Design an Interview Schedule (about a page of interview questions).
[Come to the class on Wed Dec 3rd with ideas for your research question and possible interview questions ready to discuss]
5) Recruit your participants, considering how to describe your study and ethical implications of the exercise.
6) Pilot your interview schedule, assess how things went, may redesign some questions
7) Each student to conduct at least one interview, tape recording with the respondent’s permission.
8) Transcribe the taped interview as close to verbatim as possible
[You should aim to do all of this before the Xmas break]
9)Conduct a basic Thematic Analysis of your interview transcript(s), amending this depending on whether you are using Grounded Theory, Discourse Analysis or Narrative Analysis as your chosen method of analysing your data