|
1
|
- Claudine BURTON-JEANGROS
- Department of sociology
- University of Geneva
- Training Course in Reproductive Health/
Sexual Health Research
- Geneva 2006
|
|
2
|
Plan<=
/div>
- Social sciences methods
- The research process
- Quantitative versus qualitative research
- Overview of methods
- Questionnaires
- Interviews
- Focus groups
- Participant observation
- Document analysis
- Importance of triangulation:
- example of the AIDS prevention research
|
|
3
|
- Types of research
- Fundamental research
- Applied research
- Social sciences research
- Social reality, immersion of researcher
- Abstract notions
- Objectives of research
|
|
4
|
|
|
5
|
|
|
6
|
- What is the period under study?
- What is the geographic context under study?
- General description or specific social categories ?
- What is the unit of analysis ?
- On which aspects/dimensions will the project be focused on ?
|
|
7
|
- Quantitative
- Focus on numbers, statistical analysis
- Questionnaires
- Secondary analysis of collected data
- Qualitative
- Focus on words, processes, subjective definitions
- Qualitative interviews
- Focus groups
- Participants observation
|
|
8
|
- Operationalisation process
- Concepts
- Dimensions
- Indicators
- Questions
|
|
9
|
- Types of questions
- Socio-demographic questions, personal data
- (age, sex, education, reli=
gion,
)
- Behaviours, practices
- Knowledge, cognition
- Attitudes, opinions, values, representations
|
|
10
|
- Format of questions
- Closed questions
- Open-ended questions
- Semi-open questions
- Vignettes, scenarii
|
|
11
|
- Formulation of questions
- Short questions (max 20 words)
- Clear and precise questions
- Appropriate vocabulary
- Neutral formulation
- 1 question only per question !
|
|
12
|
- Structure of questionnaire
- Sections, order is important
- Coherence, logical flow
- Provide instructions (1 or several answers are possible,
skip/filters,
)
- Socio-demographic data at the end
- Difficult/sensitive questions: not at the beginning
- Limit length of questionnaire
|
|
13
|
- Pre-test / pilot survey
- Passation modes
- Face to face
- Postal survey
- Self-administered
- Telephone survey
- Interviewers training
|
|
14
|
- Population versus sample
- Representativity
- Sampling issues
- Random sample
- Non-random sample
- Non-respondents
|
|
15
|
|
|
16
|
- Analysis of data
- Univariate analysis: frequencies
- Bivariate analysis: crosstabulations
- Multivariate analysis: regressions, typologies,
- Focus on « WHY ? »
- But difficulties with causality (correlations)
- Several softwares: epi-info, spss, sas,
|
|
17
|
- Several uses
- Main method
- Explore and prepare a questionnaire survey
- Supplement a questionnaire survey or participant observation, analy=
sis
of documents,
|
|
18
|
- Focus on :
- subjective representations, discourses
- processus, mechanisms
- Focus on « HOW ? »
|
|
19
|
- Research definition, hypotheses
- Interview grid
- Weak structure: one main topic and possible dimensions (discovery
approach)
- Structured grid: list of topics, hierarchical order
|
|
20
|
- Sampling
- Categories / groups (age, sex, social class, urban-rural, experienc=
e of
a specific event,
): allow for comparisons
- Size: smaller than for questionnaire survey
- Criteria
- Diversification versus redundancy
- Time, resources
|
|
21
|
- Access to interviewees
- Direct access: existing files, list, door-to-door
- Indirect access: thru a third party
- Snow-balling
|
|
22
|
- Interviewing situation
- Environment (timing, private or public space, work or home environm=
ent)
- Communication: information on the research project, on selection of
interviewees)
- Recording and confidentiality
- Role of interviewer
|
|
23
|
- Analysis
- Transcription of recording
- Interview analysis (processes)
- Thematic analysis (analysis grid, coding of texts)
- Specific software (atlas-ti, nudist,
)
|
|
24
|
- Origins: media studies, marketing
- More interviewees
- Easier access to individual discourses
- Gain of time and resources
|
|
25
|
- Natural group or artificial group?
- Social homogeneity of participants
- 5 to 10 participants
- 3 to
? groups
|
|
26
|
- Environment
- Duration: 1 to 2 hours
- Important role of moderator
- Recording
- Short questionnaire for personal data
- Transcription
- Analysis
|
|
27
|
- Overt observation
- Covert observation
- Public/open spaces
- Private/closed spaces
|
|
28
|
- Types of data:
- Setting: place, characteristics of space, decor,
- People: who, positions of people,
- Actions: what is happening?
- Discourse: what is being said?
|
|
29
|
|
|
30
|
- « You see, but you do not observe.
The distinction is clear. For example, you have frequently se=
en
the steps which lead up from the hall to this room »
- « Frequently »
- « How often? »
- « Well, some hundreds of times »
- « Then how many are there »
- « How many? I dont know »
- « Quite so! You have not observed. And yet you have seen. That=
is
just my point. Now, I know there are seventeen steps , because I have
both seen and observed »
-
(In Sir Arthur Conan Doyle, "The Adventures of Sherlock
Holmes. A Scandal in Bohemia")
|
|
31
|
- Anonymity and confidentiality
- Informed consent
- Incentives to respondents?
|
|
32
|
- 1980s: risks of an aids epidemic?
- few data on sexual behaviours
- sensitive research topic
- Difficulties for funding, political barriers
|
|
33
|
- Quantitative surveys by the late 1980s
- Theoretical approach: individual focus, education, rational actor
- Data collected on:
- Age at first sexual intercourse
- Sexual behaviours (risk behaviours, condom use, behaviour change)=
li>
- Number of partners
- Impact of prevention: knowledge, behaviour change
|
|
34
|
-
men and women count differently. That is to say they do not
define their partners in the same way: women would tend to consider =
only
partners that really mattered to them and therefore they would
under-estimate the total number of their partners; whereas men would
consider that reporting a large number of partners prove their viril=
ity
and therefore they would tend to over-estimate the number of their
partners and/or include partners that they only kissed or fondled&nb=
sp;
(Lιridon et al. In Hubert et al. 1998)
|
|
35
|
- 2. Relational approach: qualitative interviews (1990s)
- Theoretical approach : values, interactions, gender dimension
- Data collected on
- Men and women expectations, values
- Power relationships
- Communication between partners
|
|
36
|
- A choice of methods
- Importance of research design, research as a process
- Organize work: stages, schedule (ie according to time and resources)=
|