Ethnographer’s Toolkit

Christopher E Strauss - created 23 November 2000, edited 27 November 2000

 

Warning: the outlined notes below contain some paraphrasing, but for the most part are directly excerpted from the source articles “as is,” often verbatim, for simplicity and to avoid problems with interpretation later.  Text lifted from these notes will normally need to be quoted and referenced by citation.

LeCompte, M.D., & Schensul, J.J. (1999).  Designing and conducting ethnographic research.  Walnut Creek, CA:  AltaMira Press.

v     A fourth characteristic of ethnographic research is that it uses inductive, interactive, and recursive processes to build theories to explain behavior and beliefs under study p.15

v     Glaser and Strauss (1967) refer to this process as “grounded theory.”  Spradley (1979) refers to the same process as domain and structural analysis, whereas LeCompte and Preissle (1993) and Merriman (1988) refer to it as recursive analysis.

v     (Excellent table of the characteristics of major research paradigms on page 59-60)

v     Research Paradigm Chap.3 pp.41-60

§         NOT Positivistic or Critical, or Interpretive (Phenomenology)

§         Primary:  Interpretive Approach (or Constructivist) p.48

·        Social construction of reality

·        …culture as both cognitive and affective, as reflected in shared meanings and as expressed in common language, symbols, and other modes of communications.

·        …always defines shared constructs and meanings as “situated”

·        … approaches are inherently participatory because meaning can be created only through interaction.

·        Interpretivists believe that cultural beliefs and meanings are:

¨      Socially constructed

¨      Situated, and therefore relative to a specific context

¨      Not fixed

¨      Negotiated

¨      Multiply-voiced

¨      participatory

§         Supporting:  Ecological p.51

·        View individuals as functioning in a social context that influences their behaviors. (distributed cognition?)

·        See these levels, institutions, or sectors within a community as systematically related to and affecting one another. (networks)

·        Think that research that is guided by the ecological paradigm should identify those contextual elements with the greatest influence on individual or institutional behavior. (incentives to create knowledge)

§         Supporting:  Social Network p.52

·        … combines the work of a number of people:

¨      Communication specialists, who are concerned about the flow and exchange of information in communities…

¨      Resource specialists … who are interested in the ways in which … organizations relate to one another to serve clients.

·        Social network researchers conduct research with social networks in several different ways:

¨      Through ethnographic mapping or description…

¨      Through survey techniques…

¨      Through “snowball” or network sampling…

·        Network research is one important component in an ecological approach and can be incorporated readily into the work of all others described in this chapter.

§         (Yes, this is an ethnographic study)  Ethnographies are case studies because of their focus on a single entity, but they differ from case studies in general in that … they always include in their focus the culture of the group or entity under study p.84

§         Population p.110 (from Chapter 5 – Choosing and designing an ethnographic research project)

·        Populations are usually made up of human beings, but they can also constitute communities, organizations, programs, animals, places, things, time periods, documents, words, phrases, sentences or paragraphs in interview texts and transcripts, specific activities or bits of behavior, and other such units.

§         Theoretical sampling, or selecting for conceptual considerations p.113

·        Criterion-based selection … involves choosing study participants because they possess characteristics related to the study’s central questions

¨      Extreme or dichotomous case selection … those representing the ends of a defined population continuum

¨      Typical case selection … based on a known average for the population

¨      Unique case selection … a nonreplicable event or situation

¨      Reputational case selection … based on recommendations by experts

¨      Bellwether or ideal case selection … choosing a case because it possesses all of the necessary components for program success or maximum presence of characteristics of interest to the researcher

¨      Comparable case selection … because of their similarity to central characteristics of interest to the researchers

§         Defining, Operationalizing, and Bounding a Population p.115

·        To obtain a population that can actually be studied, researchers must “operationally define” and then “bound” the population.

·        Naturally bounded (islands, high-risk sites, enrolled, self-defined)

·        Artificially bounded (share a common characteristic but are part of general population)

·        Those to be included in the study are the “units” to be studied p.118

§         What is a unit of analysis?

·        … researchers need to define units of analysis in operational terms because they need to be able to identify discrete individuals (or units) from the given population for observation, questioning, and/or counting.  … some simple rules for defining units of analysis:

¨      They must be countable

¨      They must be measurable or describable

¨      They must be locatable

¨      Their beginnings and endings must be identified (i.e., the researcher must be able to distinguish clearly one unit of analysis form another)

¨      If the researcher is planning a survey, units of analysis also must be able to be located on a list of individuals or able to be enumerated or counted so that a list can be created and a sample drawn.

Ø      Collecting Ethnographic Data (Chapter 6) p.127

§         Data Collection Methods

·        Observation

·        Tests and repeated measures

·        Population or sample survey

·        Ethnographic interview

·        Content analysis of secondary text or visual data

·        Focus group interviews

·        Elicitation methods

·        Audiovisual methods

·        Spatial mapping

·        Network research

§         Using multiple or alternative sources of data: Triangulation and redundancy p.130

·        … multiple sources of data serve as sources of confirmation or corroboration for each other

·        … multiple sources of data … make sure that if one data set or source proves to be unreliable or incomplete, others will suffice to provide the information needed to answer each research question posed

·        Researchers call the process of creating redundancy, triangulation (Denzin, 1978).

·        Components of a data collection matrix p.138

¨      Which research questions are to be asked

¨      Which data will answer those questions

¨      Where, and from whom, those data can be obtained

¨      In what form the data will be collected

¨      Who will be responsible for collecting, analyzing, and writing up the data

¨      When each stage of data collection, analysis, and report writing will begin and end

¨      How, by whom, and to whom results will be disseminated

Ø      Data analysis: how ethnographers make sense of their data p.147

§         … ethnographers begin data analysis well before data collection is complete – even as early as when the first few interviews have been conducted

§         Ethnographers need to engage in several levels of analysis as they go along, because doing so helps them to make sense of what they are observing.

§         … patterns actually emerge because the researcher is engaged in a systematic cognitive process that takes place in three stages.  These stages may be termed “item,” “pattern,” and “constitutive” or “structural” analysis…

§         The initial taxonomy or arrangement of items, patterns (domains), and structures derived from the initial model casn be organized as hierarchies…

§         These initial conceptual categories are guides to observation and interviewing.  They are continuously enhanced, expanded, subdivided, and enriched throughout the course of the research until, in the end, a much more elaborated system of organizing, arranging, and eventually, coding data has emerged from the data that can be applied to the entire data set.

§         Item level of analysis p.151

·        Glaser and Strauss (1967) call this process one of “constant comparison,” in which each item – whether identified previously or just emerging – is compared to all other items so that all can be clearly identified, operationalized, and distinguished one from the other.

¨      Operationalization means defining a concept concretely in such a way that it can be understood, observed, or categorized accurately by any researcher reviewing the same data or observing the same event.

§         Pattern level of analysis p.154

·        Patterns consist of groups of items that fit together, express a particular theme, or constitute a predictable and consistent set of behaviors

§         Constitutive or structural level of analysis p.155

·        … overall structural element or theme …

Schensul, S.L., Schensul, J.J., & LeCompte, M.D. (1999).  Essential ethnographic methods: observations, interviews, and questionnaires.  Walnut Creek, CA:  AltaMira Press.

v     (Mostly narrative and discussion of how to plan and perform each type of data collection)

v     Building formative theoretical models (Chapter 2) p.9

Ø      Creating an initial focus calls for selecting one or more paradigms to organize the inquiry and for building a formative theory.

v     Continuum of levels of abstraction p.14 (from this series as well as Pelo & Pelto, 1978 pp. 9-13)

Ø      General level and models

§         Paradigms: explain and interpret relationships among structures

Ø      Middle-range theory and models

§         Structures: middle-range theory explains and interprets structures or the relationships among patterns

Ø      Low-order propositions

§         Patterns: lower-order propositions explain and interpret patterns or the relationships among units

Ø      Modes of observation

§         Units: Modes of observation operationalized units or the relationships among facts

Ø      Facts: Real-world modes, items

v     Operationalization (Chapter 3)

Ø      … measurement … does not require mechanical devices, and … many measurements are not expressed in terms of scores or numbers.  Measurement simply permits the researchers to assess the variation in a particular variable p.50

Ø      Operationalization processes

§         Definition: establishing boundaries to distinguish this particular phenomenon from others and to communicate that distinction to other researchers

§         Measurement: describing or measuring the phenomena

§         Establishing validity: assessing the adequacy of the measurements in representing the phenomenon

§         Establishing reliability: ensuring consistency in the measurement process over time and among the researchers carrying out the measurement activities

Ø      Operational definitions consist of a set of instructions on how to measure a variable that has been conceptually defined p.51 (Bernard, 1995)

Ø      Operationalizing – transforming information from one level of abstraction to another p.52

§         Domain (Independent and Dependent are defined at this level and carry through to lower levels)

·        Factor

¨      Sub-Factor

Ø      Variable

§         Item / Fact

v     Exploratory or open-ended observation (Chapter 5) p.91

Ø      Participation is always defined by researcher presence at the event being observed p.92

Ø      … ethnographers usually start by doing the following:

§         Observing settings

§         Observing and tracking events and event sequences

§         Counting, census-taking, and ethnographic mapping

§         Searching for indicators of socioeconomic (hierarchical) difference

v     Using archival and secondary data (Chapter 9) p.201

Ø      Archival data are materials collected for bureaucratic, service, or administrative purposes and transformed into research data

Ø      Secondary data are data collected by other researchers for their own research purposes…

§         Local secondary data … on the population under study

§         Non-local secondary data … conducted elsewhere with comparable populations on similar topics (i.e., data collected in call center situations by Ackerman & Halverson, etc.)

Ø      No archival data fully represent the study population p.212

v     Validity and reliability in ethnographic research (Chapter 11) p.271

Ø      validity is concerned with accuracy and dependability of instruments and observations p,274

§         Internal validity refers to the correspondence between measures and the reality of the field situation - the extent to which scientific observations and measurements … authentically represent the reality in which the people studied live – as they define it …

§         Construct validity refers to whether instruments measure what they are assumed to measure - instruments such as tests, questionnaires, and interview guides, although it is also relevant to the questions researchers ask in interviews … the degree to which the questions or measures used really assess what they are assumed to measure p.275

§         External validity refers to the applicability of representations to other groups – the degree to which such representations can be applied to other groups.

Ø      Reliability is concerned with whether the research results can be replicated … using the same methods

§         Internal reliability refers to the match between the constructs and a data set – the degree to which other researchers would match …constructs with a particular data set in the same way …

§         External reliability refers to the comparability of results in repeat studies using the same methods – independent researchers would discover the same phenomena or generate the same constructs…

 

Schensul, J.J., LeCompte, M.D., Trotter, R.T.I., Cromley, E.K., & Singer, M. (1999).  Mapping social networks, spatial data, and hidden populations.  Walnut Creek, CA:  AltaMira Press.

v     Friends, relatives, and relevant others: Conducting ethnographic network studies (Only looked at chapter on mapping social networks; other two do not apply to my study) p.1

Ø      Ethnographic network mapping, a type of network research that can be used to describe … work groups ... problem-solving groups p.4

§         Ethnographic description of these networks includes the following (adapted for my study):

·        The identity of the people in them

·        How people define or self-define themselves as group members

·        The rules people use for including and excluding members

·        (work, school, and social) relationships within the groups

·        Information about how group members (accomplish their service mission)

Ø      Ego-centered or personal network research p.5

§         … the personal networks of individuals, also called “egos” or “focal individuals”

§         “Alters” are those people whom index individuals identify as members of their personal networks

§         Ego-centered or personal network research gives ethnographers the ability to describe the context of people’s lives beyond their individual social and psychological characteristics

Ø      Full relational social networks p.6

§         … a naturally occurring social network

§         May have subgroups, and bridges to other networks

§         The social network approach allows ethnographers to move beyond the level of the individual and the analysis of individual behavior in the social context where most people spend the vast majority of their lives, living and interacting with the small l groups that make up the world around them p.7

Ø      Key concepts for ethnographic approaches to social networks

§         Boundaries and Bridges

·        Boundaries constitute the edges of networks and are defined by rules for entry and exit from groups as well as by other cultural patterns of participation that differentiate one group from another p.8

·        Bounded groups (where you have identified the boundaries, the edges of networks)

§         Network typologies p.9

·        … typologies of networks or descriptions of different kinds of networks

·        Case study: three structural conditions observed p.11

¨      Openness

¨      Social bonds

¨      Social interactions

·        … yielded four distinct, internally consistent and externally divergent networks

¨      Long-term (behavior-based)

¨      Family-based

¨      Friendship-based

¨      Convenience networks

§         Interaction and “Embedded Behaviors” p.15

·        …ethnographic description … can discover behavioral details and patterns of communication and influence specific to the group, which can then be quantified when the ethnographer wishes to measure them in other networks

§         Cross-group differences p.17

·        Voluntary association are those that people join voluntarily to meet a variety of their personal, economic, and social needs

Ø      Ego-centered approaches p.18

§         Ties are links between people that are measured by perceived intimacy and frequency of association p.25

§         (all kinds of instructions on how to do data collection and analysis of egocentric network data)

Ø      Full network relationships: reciprocal network information p.30

§         Ethnographic research on entire relational networks includes information about:

·        Connectedness – the extent of reciprocal relationships among individuals in a network

·        Power and influence – the degree to which a person in a network gives and receives information or other resources

·        Role relationships – those associated with a particular place or position in the social network

·        Cliques and components – subunits within a larger network designated by the greater strength of their interconnectedness compared with other parts of the network

·        Density – the actual number of relationships found in a network compared to the total number of possible relationships

§         Communication flow (connectivity) … can be assessed by measuring the presence or absence of changers in the level of information flow with in the network

·        … can be characterized by several measures, including the amount of information that passes through a network, etc.

¨      … identification of gatekeepers

§         Distance (social cohesion) and segmentation (similar connectivity to others) p.31

§         Positions and roles of network members (centralization) p.32

·        Multiplex ties in a group means that each group member has many different relationships with most or all other members in a group (classic communications configuration) p.38

Ø      Network sampling strategies

§         … all individuals in the full network

§         … a randomized sample of index individuals’ alters

§         … by screening and selecting individuals because of their special characteristics

§         … by choosing an index person and selecting from among all of that person’s alters (random walk design)