reading 11 - interviews Flashcards
continuum of interviews
informal interviewing, unstructured interviewing, semistructured interviewing, structured interviewing
informal interviewing = total lack of structure or control
- researcher tries to remember conversations heard during the course of a day in the field
- method of choice at beginning of participant observation fieldwork
- is deceptive: keep people from knowing that you’re really at work, studying them
unstructured interviewing = formal + undeceptive: you’re sitting down for an interview + based on clear plan and minimum of control over the people’s responses
- get people to open up and let them express themselves in their own terms + own pace
- lot of ethnographic interviewing is unstructured
- appropriate when you have lots of time
semistructured interviewing = interviewing based on use interview guide: written list of questions and topics that need to be covered in a particular order = control of what you want in the interview, but leaves room to follow new leads
- in situation where you won’t get more than one chance to interview someone
- interviewer maintains discretion to follow leads, but interview guide provides set of clear instructions
- to get reliable, comparable qualitative data
structured interviewing = people are asked to respond to as nearly identical a set of stimuli as possible
- use of interview schedule: explicit set of instructions to interviewers who administer questionnaires orally
- e.g. if informant answers A, ask question 29, if they anser B, ask question 30
unstructured interviewing
used by both positivists and interpretivists
- ethnographers may use unstructured interviewing in developing structured interview schedules
useful when you want info on lived experiences + want to build rapport/trust + informants who wouldn’t tolerate a more formal interview
can be used for studying sensitive issues if you’re good at probing
getting started: assure people of anonymity and confidentiality + tell them why them + encourage informants to interrupt you with anything important + ask permission to take notes and record = be open about intensions
let the informant/respondent lead: keep the conversation focused on a topic while giving the respondent room to define the content of the discussion
- get people on to a topic of interest and get out of the way, let the informant provide info they think is important
- researcher defines the focus of the interview, the respondent the content
probing
= to stimulate a respondent to produce more information without injecting yourself so much into the interaction that you only get a reflection of yourself in the data (e.g. ask where? rather than: place A, or place B or maybe place C?)
different types of of probes:
- silent probe = remaining quiet and waiting for an informant to continue (can include a nod or uh-huh)
- sometimes this isn’t even a probe, but an appropriate cultural behavior
- high-risk technique: can become awkward, lose credibility - echo probe = repeating the last thing someone said and asking them to continue (I see …, then what happeneed)
- don’t use it too often: people will be annoyed you keep repeating what they say - uh-huh probe = make affirmative comments (uh-uh, right, yes)
- is a neutral probe - tell-me-more probe: can you tell me more about that? why exactly do you say/feel that?
- don’t use it too often - long question probe: use grand tour question to get the interview flowing, focus on really specific questions later
-> more response (not necessarily better)
- with sensitive questions, start with run-up, keep them short and follow with more questions - probing by leading: directive probes: all questions are in a way leading, learn to do it well
- baiting: the phased-assertion probe = act like you already know something to get people to open up, this way you can get confirmation
respondents can be verbal, in this case you need to learn to cut them of gracefully to stay on topic
some respondents may be nonverbal: i don’t know
4 types of idk: i don’t know and don’t care +i don’t know and it’s not your business + i don’t know, actually I do but you’re not interested in what I have to say + i don’t know and which you’d change the subject bc i’m uncomfortable
+ i wish i could help you but i really don’t know
learning to interview
practice + have an experienced interviewer monitor/criticize you
learn from yourself: pay attention to your words (only ask questions you can defend when people ask why you ask that question
importance of language: practice/learn to interview in a certain language before actually going out in the field
pacing the study:
two of the biggest issues researchers who do semistructured interviews = boredom and fatigue: asking the same questions over and over
-> boredom makes interviews shorter
-> spread the project out if possible
- tradeoff: the longer a project takes, the less likely that the first interviews and the last interviews will be valid indicators of the same things -> maybe interesting to go back to your first interviewees after a while
presentation of self
cordial-but-nonjudgmental
- try to probe without injecting your feelings into the interview
- some situations to painful -> impossible to maintain neutral facade
if someone tells: i hate watching all my friends die one at a time, it might be weird to uh-huh
little things mean a lot
- dress appropriately: use common sense
- venue matters: some places are wrong for interviews (use common sense)
recording interviews
don’t just rely on your memory, use a voice recorder
- when people feel uncomfortable: keep it off at the beginning, start with small talk OR give people a mic with on/off switch (people rarely use it + gives sense of controle) OR turn it off temporarily
equipment: have more recorders than you think you need, upload data regularly, use separate microphone for better quality, omnidirectional mics are nice (allows interviewee to walk around whilst answering), test before using, start with fresh batteries
transcribers and VR software: doing it yourself costs 6-8 hours pr hour of recording
- transcription software = lets you control the recorder (start, stop, move forward and backward) with keyboard
- transcriber machines = lets you control the recorder with foot pedal
- VR software = voice recognition: listen to interview and repeat the words -> software notes it down
!recording is not a substitute for taking notes: did they seem nervous? how much probing did you have to do?
- ask before taking notes, some interviewees seem offended by it
focus groups
complement surveys or individual interviews (don’t replace them)
widely used to find out why people feel as they do about something or the steps people go through in making decisions
usually 6-12 members per group + moderator
focus group research is based on a series of focus groups (not based a single) + groups are homogeneous on certain IVs (just like in experimental and sampling design)
are focus groups valid?
focus groups should be used for collection of data about content and process (bc they offer more detail), not for estimating population parameters of personal attributes
- for that you need a method that produces numbers, like surveys
value of focus groups = produce ethnographically rich data (if captured with transcription)
running a focus group
moderator gets people talking about an issue + needs to silence some and get others to talk more
- break the ice: get some to talk, others will follow
- create a supportive + nonjudgmental sfeertje
- don’t lead too much, don’t put words in people’s mouths
analyzing data from focus groups: formal content analysis or qualitative analysis
- its best to record focus groups (-> try to prevent people from speaking at the same time)
- often two staff members: one moderator + one that jots down name of each person who speaks and the first few words they say -> makes it easier for transcriber to identify the voices
!!not all group interviews are focus group interviews: in fieldwork it can happen that people insert themselves in a conversation
response effects
= measurable differences in the responses of people being interviewed that are affected by:
- characteristics of the interviewers and those being interviewed
- features of the environment where the interview takes place (e.g. presence of a third party vs private)
- nature of the task that people are asked to perform
- mode of the interview (face-to-face, telephone, internet)
e.g.
- 1929 Rice: political orientation of interviewers can affect what they report people told them
- 1942 Katz: middle-class interviewers got more politically conservative answers in general from lower-class respondents than did lower-class interviewers
- Robinson and Rhode: interviewers who looked non-Jewish + had non-Jewish-sounding names 4x likely to get anti-Semitic answers to questions about Jews
gender-of-interviewer effects = women are better than men as interviewers: fewer don’t know responses + more responses
- but it appears educated respondents of both sexes shift their answers toward the socially desirable positions they think are held by female interviewers
- women more open about if they had sex to male interviewers
race of interviewer effect =
1958 Lenski and Leggett: African Americans 4x more likely than whites to agree to anything (even contradictory statements) bc the interviewers were almost all white and of higher-perceived status (this effect was not ended with election Obama)
- example of deference effect
language and culture of the interviewer = when you have multiple interviewers, keep track of the gender, race and ethnicity of them and test for response effects
- identify sources of bias
social desirability effect = when people tell you what they think will make them look good according to prevailing standards of behavior and thought
- effect is influenced by how you ask the question (e.g. ask what did you do this weekend rather than did you go to church)
third party-present effect: many interviewers are not entirely private: when 3d party is present people are less likely to tell/report some things (e.g. thoughts about suicide) or more likely to report something (e.g. when partner in the room report more marital conflict)
- suspection: people own up to more sensitive things when their partner is in the room and it will be obvious to them that they’re lying
threatening questions: disclosure of information about socially undesirable behavior increases with perception of anonimity (so e.g. more open in self-adminsitered than face-to-face surveys)
- asking about other people increases reports about socially undesirable behavior
- when you give people choices that include a big number of any behavior, you get higher responses (e.g. if you ask people if they had sex with 0,1,2,3,4,5,6,7,8 people last week)
!!asking interviewers to record interviews produces higher response: interviewers know their work can be scrutinized -> probe more
expectancy effect
tendency for experimenters to obtain results they expect, not simply because they have correctly anticipated nature’s response, but rather bc they have helped to shape that response through their expectations
(strictly speaking not a response effect)
behavior of the researcher can affect the behavior of informants
confirmation bias
= tendency to see what we expect to see and for any new evidence we uncover to confirm rather than challenge what we already believe
deference/acquiescence effect
= when people tell you what they think you want to hear bc they see you as being of higher status
accuracy
you have to take the word of people
- dietary recall: self-reports of what people ate often inaccurate (test with DLW technique: special water, test urine to see intake of food)
- 1961 census in Ethiopia -> 23% of women underreported the nr of children (didn’t count babies who die before reaching age of 2)
La Pierre Discovers the problem:
1930-32: traveled with Chinese couple through the US, were not denied service in any restaurant + only in one hotel
after experiment ended, La Pierre sent questionnaire to each of the 250 establishments they visited: 92% answered they would not accept members of the Chinese race as guests
- some problems with this research: self-adminstered questionnaire (respondents may not be same persons who gave service) + survey didn’t mention the couple would be accompanied by a white man
why are people inaccurate reporters of their own behavior?
- Once people agree to be interviewed, they have a personal stake in the process and usually try to answer all your questions—whether they understand what you’re after or not.
- Human memory is fragile, although it’s clearly easier to remember some things than others.
- easy to remember rare event, otherwise people use estimation rules - interviews are social encounters, people manipulate them to what they think is in their advantage (boys tend to exaggerate, girls to minimize reports on sexual experience)
- people can’t count a lot of behaviors -> use rules of inference (report what must have happened rather than what actually happened)
aided recall
landmarks = effective ways to jog informants’ memories about details
- try to establish personal milestones for each informant + ask them to report on what has happened around and after then
aided recall increases nr of events recalled, but increases forward telescoping (people report that something happened e.g. one month ago while it actually happened 2 months ago)
- backward telescoping is rare
- can correct for telescoping when you interview people more than once by reminding people what they responded last time, ask how often something happened since then
event history + life history calanders = produce a lot of landmarks + are effective aids to recall
- helps tie life events to one another and help people remember details (e.g. not just i changed jobs, i was really said bc … and changing jobs helped to keep my mind off it)
