Qualitative Research Qualitative Research MethodsMethods

Mary H. Hayden, PhDMary H. Hayden, PhDNCAR Summer WAS*IS NCAR Summer WAS*IS

July 20, 2006July 20, 2006

Presentation OutlinePresentation Outline

Distinguishing qualitative and quantitative Distinguishing qualitative and quantitative approachesapproaches

Qualitative methods – Qualitative methods – Types of qualitative methodsTypes of qualitative methods Advantages vs. DisadvantagesAdvantages vs. Disadvantages

Real World ExamplesReal World Examples

Qualitative and Quantitative ApproachesQualitative and Quantitative ApproachesQualitative Quantitative

(Usually) Non-probability based sample

Typically a probability-based sample

Non-generalizable Generalizable

Answers Why? How? Answers How many? When? Where?

Formative, earlier phases Tests hypotheses, latter phases

Data are “rich” and time-consuming to analyze

Data are more efficient, but may miss contextual detail

Design may emerge as study unfolds

Design decided in advance

Researcher IS the instrument Various tools, instruments employed

Qualitative “vs.” Quantitative?Qualitative “vs.” Quantitative?

Research question guides the choice Research question guides the choice Choice of methods is situationalChoice of methods is situational ComplementaryComplementary

Research more complete with bothResearch more complete with both Each requires resources and timeEach requires resources and time

Can be amended for rapid analysisCan be amended for rapid analysis

Types of qualitative methodsTypes of qualitative methods

InterviewsInterviews

InformalInformal – researcher is required to recollect – researcher is required to recollect discussiondiscussion

UnstructuredUnstructured – e.g. ethnographic interviewing – – e.g. ethnographic interviewing – researcher allows interview to proceed at researcher allows interview to proceed at respondent’s pace and subjects to vary by respondent’s pace and subjects to vary by interviewee (to an extent)interviewee (to an extent)

Semi-structuredSemi-structured – researcher uses an interview – researcher uses an interview guide guide

StructuredStructured – researcher uses identical stimuli – researcher uses identical stimuli and adheres to interview scheduleand adheres to interview schedule

Semi-structured Semi-structured InterviewsInterviews

Focus GroupsFocus Groups

Recruited to discuss particular topicRecruited to discuss particular topic One focus group is ONE unit of analysisOne focus group is ONE unit of analysis Complement surveys – often the 1st step Complement surveys – often the 1st step

in tapping critical questions to be used in a in tapping critical questions to be used in a surveysurvey

Identify why people feel certain way and Identify why people feel certain way and elucidate steps in their decision-making elucidate steps in their decision-making processprocess

Focus Group MethodsFocus Group Methods

Ideal size: 6 – 12 people and a Ideal size: 6 – 12 people and a moderator/note takermoderator/note taker

Series of groups is necessary for validitySeries of groups is necessary for validity Homogeneity and anonymity in selection Homogeneity and anonymity in selection

of groupsof groups people may open up with others who are people may open up with others who are

perceived to think along similar lines perceived to think along similar lines ANDAND whom they may never see againwhom they may never see again

Focus Group Methods, cont.Focus Group Methods, cont.

Often segment according to expected Often segment according to expected meaningful differences (e.g. disease meaningful differences (e.g. disease status, gender…)status, gender…)

Running a focus group – fine line between Running a focus group – fine line between leading too much and not getting people to leading too much and not getting people to contributecontribute

Important to keep discussion on topic w/o Important to keep discussion on topic w/o shutting people downshutting people down

No right or wrong answersNo right or wrong answers

Focus Group Methods, cont.Focus Group Methods, cont.

Coding/AnalyzingCoding/Analyzing

Tapes are usually transcribed verbatimTapes are usually transcribed verbatim Text is sorted into emergent themes by at Text is sorted into emergent themes by at

least 2 researchers to ensure validity using least 2 researchers to ensure validity using pile-sort method or computerized version pile-sort method or computerized version such as CDC’s EasyText (free!)such as CDC’s EasyText (free!)

Themes are compared with field notes Themes are compared with field notes taken by second researchertaken by second researcher

Focus Group Methods, cont.Focus Group Methods, cont.

Thematic coding factors:Thematic coding factors:

1. Frequency – number of times something 1. Frequency – number of times something is mentionedis mentioned

2. Specificity – details2. Specificity – details

3. Emotion – enthusiasm, passion, etc. in 3. Emotion – enthusiasm, passion, etc. in responsesresponses

4. Extensiveness – how many different 4. Extensiveness – how many different people said somethingpeople said something

Structured InterviewsStructured Interviews

QuestionnairesQuestionnaires

Three methods:Three methods:

1.1. Face-to-face interviewsFace-to-face interviews

2.2. Self-administered questionnairesSelf-administered questionnaires

3.3. Telephone interviewsTelephone interviews

Face-to-Face Interviews (and Face-to-Face Interviews (and intercept interviews)intercept interviews)

Advantages:Advantages: Can be used with respondents who wouldn’t be Can be used with respondents who wouldn’t be

able to provide information in another format – able to provide information in another format – bedridden, illiterate, etc.bedridden, illiterate, etc.

Researcher can elicit more in-depth response or Researcher can elicit more in-depth response or fill in information if participant doesn’t fill in information if participant doesn’t understand the questionunderstand the question

Different data collection techniques – open-Different data collection techniques – open-ended questions, visual aids, etc.ended questions, visual aids, etc.

Certainty about who answered the questionsCertainty about who answered the questions

Face-to-Face Interviews, cont.Face-to-Face Interviews, cont.

Disadvantages:Disadvantages:

Intrusive and reactiveIntrusive and reactive Cost time and moneyCost time and money Difficult to locate respondents for callbacksDifficult to locate respondents for callbacks

Self Administered QuestionnairesSelf Administered Questionnaires

Advantages:Advantages: Post Office locates participantsPost Office locates participants Everybody gets the same questionsEverybody gets the same questions Researcher can ask more complex Researcher can ask more complex

questions questions No response effect (willing to divulge more No response effect (willing to divulge more

info w/o face-to-face contact; less likely to info w/o face-to-face contact; less likely to try to impress interviewertry to impress interviewer

Can be computer-basedCan be computer-based

Self Administered Questionnaires, Self Administered Questionnaires, cont.cont.

Disadvantages:Disadvantages: No control over participant interpretationNo control over participant interpretation Low response ratesLow response rates Uncertainty about who actually filled out the Uncertainty about who actually filled out the

questionnairequestionnaire Useless with non-literate, illiterate Useless with non-literate, illiterate

populations (same problem with English populations (same problem with English language sampling) or hard-to-reach language sampling) or hard-to-reach populationspopulations

Telephone InterviewsTelephone Interviews

Advantages:Advantages: Combo of face-to-face personal quality Combo of face-to-face personal quality

with impersonal self-administered with impersonal self-administered questionnairesquestionnaires

Inexpensive and convenient (maybe)Inexpensive and convenient (maybe) Safe for interviewersSafe for interviewers

Telephone Interviews, cont.Telephone Interviews, cont.

Disadvantages:Disadvantages:

Changing demographics – more cell phones? Changing demographics – more cell phones? May miss certain population segmentsMay miss certain population segments

Survey must be short or people will hang upSurvey must be short or people will hang up ““No Call Lists” presenting increasing challengeNo Call Lists” presenting increasing challenge

Focus Groups vs. SurveysFocus Groups vs. Surveys

Surveys offer quantitative measurements Surveys offer quantitative measurements based on a representative samplebased on a representative sample

Focus groups offer content insight – the Focus groups offer content insight – the whywhy of what people think of what people think

Real World ExamplesReal World Examples West Nile Virus (WNV) focus group research in West Nile Virus (WNV) focus group research in

CO in 2003 CO in 2003 11 semi-structured focus group interviews were 11 semi-structured focus group interviews were

conductedconducted Groups were recruited through community Groups were recruited through community

gatekeepers and composed of participants from gatekeepers and composed of participants from senior centers, health departments, neighborhood senior centers, health departments, neighborhood and volunteer organizations and local businessesand volunteer organizations and local businesses

Purpose was to explore and contextualize Purpose was to explore and contextualize use/non-use of repellent and KAP around WNV use/non-use of repellent and KAP around WNV and mosquitoes in COand mosquitoes in CO

Real World ExamplesReal World Examples

WNV in Colorado – Selected resultsWNV in Colorado – Selected results CO residents didn’t perceive mosquitoes to be a CO residents didn’t perceive mosquitoes to be a

problem in the “high plains”problem in the “high plains” The high proportion of WNV Fever cases diagnosed The high proportion of WNV Fever cases diagnosed

in some areas led to perception that “everyone was in some areas led to perception that “everyone was sick” and higher levels of concernsick” and higher levels of concern

Younger people wanted to get infected before risk of Younger people wanted to get infected before risk of neuroinvasive disease increasedneuroinvasive disease increased

Residents perceived that media overplayed the issueResidents perceived that media overplayed the issue

Real World ExamplesReal World Examples Border ResearchBorder Research

WNV focus group research in Imperial Valley WNV focus group research in Imperial Valley along MX/CA border in 2004along MX/CA border in 2004

WNV focus group research in Tijuana/San WNV focus group research in Tijuana/San Diego in 2005Diego in 2005

Traditional surveys would be especially Traditional surveys would be especially difficult due to international agreements, lack difficult due to international agreements, lack of phones, lack of sampling framework, of phones, lack of sampling framework, inability to contextualize responses with inability to contextualize responses with neighborhoodsneighborhoods

MethodsMethods Topics CoveredTopics Covered

Knowledge about WNVKnowledge about WNV Sources of informationSources of information Perceptions re presence of mosquitoesPerceptions re presence of mosquitoes Actions to avoid/kill mosquitoes/prevent bitesActions to avoid/kill mosquitoes/prevent bites

• Repellent, chemicals, household controlRepellent, chemicals, household control Existence of health committees or health Existence of health committees or health

information in local areainformation in local area

ResultsResults

Household control of breeding sites - Most Household control of breeding sites - Most people familiar with “control of standing people familiar with “control of standing water”water” Varied understandings of “standing water”Varied understandings of “standing water”

Many Many CulexCulex breeding sites beyond means breeding sites beyond means of household control—related to irrigation of household control—related to irrigation and/or sewage canalsand/or sewage canals

Results—Information Results—Information SourcesSources

Media most frequently cited as source by Media most frequently cited as source by those who knew of WNVthose who knew of WNV

Spanish speakers in US may not be Spanish speakers in US may not be accessing local newsaccessing local news Many report primarily watching satellite Many report primarily watching satellite

channels from Mexicochannels from Mexico Imperial County (for example) challenged Imperial County (for example) challenged

by lack of local stationsby lack of local stations

Results—Special PopulationsResults—Special Populations

Needs and risks among agricultural Needs and risks among agricultural workers (US) could be significant during workers (US) could be significant during an outbreak, especially among an outbreak, especially among undocumented workers if unlikely to seek undocumented workers if unlikely to seek medical caremedical care

Need approaches for non-English and Need approaches for non-English and non-Spanish speakersnon-Spanish speakers

Migrant laborers unlikely to have access to Migrant laborers unlikely to have access to television, mail, etc.television, mail, etc.

Sample Survey Sample Survey Questionnaire CritiqueQuestionnaire CritiqueWhile surveys are a strong and popular While surveys are a strong and popular

method, the following are some examples of method, the following are some examples of questions where a closed response may questions where a closed response may

present significant limitations present significant limitations

Katrina Questionnaire CritiqueKatrina Questionnaire Critique

Tell me all the places you saw or heard Tell me all the places you saw or heard information about the hurricane before information about the hurricane before Katrina.Katrina.

Environmental cues (look outside)Environmental cues (look outside)

Family or friendsFamily or friends The Weather ChannelThe Weather Channel InternetInternet Local radio stationsLocal radio stations Local television stationsLocal television stations NOAA Weather RadioNOAA Weather Radio

Cell PhoneCell Phone Other (please describe)Other (please describe)

____________________________________________________________________________________

Katrina Critique, cont.Katrina Critique, cont.

Before Katrina, how serious did you Before Katrina, how serious did you think the hurricane and its think the hurricane and its consequences would be?consequences would be?

□ □ Very serious □ Moderately serious Very serious □ Moderately serious

□ □ A little serious □ Not serious at allA little serious □ Not serious at all

Katrina Critique, cont.Katrina Critique, cont.

Do you think that you will stay in this Do you think that you will stay in this area (shelter community)?area (shelter community)?

□ □ YES □ NOYES □ NO

Katrina Critique, cont.Katrina Critique, cont.

Should this shelter need to close would Should this shelter need to close would you be willing to move another shelter?you be willing to move another shelter?

□ □ YES □ NOYES □ NO

Dengue Questionnaire CritiqueDengue Questionnaire Critique

Do you travel back and forth between Do you travel back and forth between the US and Mexico?the US and Mexico?

□ □ YES □ NOYES □ NO

If so, number of days per year?If so, number of days per year?

Dengue Critique, cont.Dengue Critique, cont.

Do you have piped water 24 hours/day?Do you have piped water 24 hours/day?

□ □ YES □ NOYES □ NO

If no, do you store water in an open If no, do you store water in an open container?container?

□ □ YES □ NOYES □ NO

Dengue Critique, cont.Dengue Critique, cont.

Do you have screens on all windows Do you have screens on all windows and doors?and doors?

□ □ YES □ NOYES □ NO

SummarySummary

Qualitative data gathering is a rich and Qualitative data gathering is a rich and important tool in some settingsimportant tool in some settings

Variety of methods availableVariety of methods available May overcome limitation of closed survey May overcome limitation of closed survey

questions and assist in development of questions and assist in development of better instrumentsbetter instruments

ReferencesReferences

Bernard, HR. (2002) Bernard, HR. (2002) Research Methods in Research Methods in Anthropology: Qualitative and Quantitative Anthropology: Qualitative and Quantitative Approaches. Approaches. Walnut Creek: AltaMira PressWalnut Creek: AltaMira Press

Krueger, RA and Casey, MA. (2000). Krueger, RA and Casey, MA. (2000). Focus Focus Groups: A Practical Guide for Applied Research.Groups: A Practical Guide for Applied Research. Thousand Oaks: Sage PublicationsThousand Oaks: Sage Publications

Puchta, C. and Potter, J. (2004) Puchta, C. and Potter, J. (2004) Focus Group Focus Group Practice. Practice. London: Sage PublicationsLondon: Sage Publications

Zielinski-Gutierrez EC and Hayden MH.Zielinski-Gutierrez EC and Hayden MH. (2006) A (2006) A model for defining West Nile virus risk model for defining West Nile virus risk perception based on ecology and proximity.perception based on ecology and proximity. EcoHealth 3(1)EcoHealth 3(1): 28-34: 28-34