DESIGNING COMMUNITY-BASED MIXED METHODS RESEARCHsites.bu.edu/miccr/files/2017/03/Badiee-2012-Designing-Community... · designing community-based mixed methods research ... conducting

41

3DESIGNING COMMUNITY-BASED

MIXED METHODS RESEARCH

MANIJEH BADIEE, SHERRY C. WANG, AND JOHN W. CRESWELL

Because we are rapidly becoming a more multicultural society, the impor-tance of research on diverse groups cannot be overestimated (Sue & Sue,2003). Hence, the study of diverse groups should avoid approaches that con-tribute to their marginalization. Moreover, the use of purely quantitativeapproaches, in which the relationship between the researcher and participantsis distant, can isolate ethnocultural communities (Jacklin & Kinoshameg,2008). In contrast, community-based research illustrates a useful approach forstudying ethnocultural populations because of the close collaboration betweenresearchers and participants (Radda, Schensul, Disch, Levy, & Reyes, 2003).Such collaboration can flow among many stages of research, such as getting toknow the community, planning and implementing the process, and evaluat-ing the impact of a program on the community (Stoecker, 2005).

Community-based studies have incorporated both quantitative and qual-itative methodologies. As such, community-based research holds the potentialfor being framed as mixed methods research and, even more specifically,community-based mixed methods research (CBMMR). Such framing meansthat systematic procedures exist for integrating the qualitative and quantitativedata, for designs to plan the projects, and for providing detailed procedures in

12812-04_CH03_4thPgs.qxd 2/17/12 1:46 PM Page 41

http://dx.doi.org/10.1037/13742-003Qualitative Strategies for Ethnocultural Research, edited by D. K. Nagata, L.Kohn-Wood, and L. A. SuzukiCopyright © 2012 American Psychological Association. All rights reserved.

Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

conducting the studies. The benefits of using multiple methods in community-based research have been explored by previous scholars (e.g., Kidder & Fine,1987; Sirin & Fine, 2008; Strand, Marullo, Cutforth, Stoecker, & Donohue,2003). However, no systematic approach to integrating community-basedresearch and the most up-to-date mixed methods research was found. Thischapter seeks to serve as a practical template for such integration.

We first review community-based research and note models being usedin this form of inquiry. Then we introduce mixed methods research notingprocedures for integrating qualitative and quantitative data and several pop-ular designs used in this integration. Next, we advance 10 illustrative empir-ical community-based studies in which both qualitative and quantitative datawere collected. As we reviewed these studies, we noted how both forms ofdata were incorporated into the studies, how mixed methods features wereincluded, and how these features were added within the stages of the Stoecker(2005) model of community-based research. We end the chapter with spe-cific recommendations for framing community-based studies as mixed meth-ods research. We state that these recommendations and the combination ofmixed methods and community-based research will enhance research withethnocultural populations and provide new tools for research.

COMMUNITY-BASED RESEARCH STUDIES

Community-based studies are often used to study ethnocultural, ethno-linguistic, and multicultural groups (Jacklin & Kinoshameg, 2008; Mosavel,Simon, van Stade, & Buchbinder, 2005; J. J. Schensul et al., 2006; S. L.Schensul, Nastasi, & Verma, 2006; Teufel-Shone, Siyuja, Watahomigie, &Irwin, 2006; Westhues et al., 2008; Woods, 2009). By definition, community-based research refers to the “process that brings researchers and communitymembers together to collaboratively conduct research on a problem of con-cern to the community” (Radda et al., 2003, p. 204). The term has also beenused interchangeably with approaches such as participatory research, actionresearch, and community-based participatory research, approaches that allshare the common purpose of addressing the unmet needs of a communitythrough research (Taylor et al., 2004). As opposed to traditional forms ofresearch, community-based studies are unique in that the emphasis is placedon the egalitarian collaboration between researchers and community mem-bers and the shared quest to address a community issue (e.g., Harris, 2006; fora more detailed understanding of the tenets of community-based research, seeIsrael, Schulz, Parker, & Becker, 1998). In the field of psychology as well ashealth sciences in general, community-based research is rooted in advocacyand social justice for populations and communities and the needs of individ-

42 BADIEE, WANG, AND CRESWELL


Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

uals who have been underserved. Thus, it is an ideal fit for understanding eth-nocultural populations that have traditionally been marginalized.

In terms of the existing community-based research studies conducted withethnocultural communities, there are limitations that must be noted. To beginwith, it is not always the case that theoretical models are used. Furthermore,when theoretical models are used to frame the research study, they havebeen discipline specific. For example, the models discussed by Mosavel et al.(2005); S. L. Schensul et al. (2006); and Woods (2009) pertained only tocommunity-based health research. Similarly, Jacklin and Kinoshameg (2008)referred to their use of a “community centered praxis” (p. 54), which was acombination of critical medical anthropology and participatory action research.

Scholars in the literature have also highlighted methodological difficul-ties and called for a need to reevaluate traditional forms of research, especiallyfor ethnocultural populations (e.g., Boynton, 2002; Keys, McMahon, Sanchez,London, & Abdul-Adil, 2004; Robins et al., 2008; Sasao & Sue, 1993; Sterk& Elifson, 2004; Taylor et al., 2004; Trimble & Fisher, 2006; Zeller, 1993).Community-based researchers have responded to the call for nontraditionalresearch by expanding beyond traditional research procedures to collect bothqualitative and quantitative data (Aronson, Wallis, O’Campo, Whitehead,& Schafer, 2007; Naylor, Wharf-Higgins, Blair, Green, & O’Connor, 2002;S. L. Schensul et al. 2006; Sirin et al., 2008; Sirin & Fine, 2007; Westhueset al., 2008; Woods, 2009). These authors often cite the many advantages tocollecting quantitative and qualitative data in community-based research(Andrew & Halcomb, 2006; Axinn & Pearce, 2006; Jacklin & Kinoshameg,2008; Linney, 2000; Sirin & Fine, 2007; Yardley & Bishop, 2008; Zeller, 1993).Strand et al. (2003) provided a powerful rationale for gathering both qualita-tive and quantitative data (and thus using mixed methods) in community-based studies:

For those used to being quantitative or qualitative researchers, [community-based research] is both and neither. In the real world, philosophicaldifferences over whether cold statistics or richly detailed stories providebetter information are irrelevant. What matters is what information isneeded to contribute to the social change effort, and this often calls formultiple methods of data collection. (p. 78)

Similarly, Sirin and Fine (2008) argued that in recent years, “working acrossmethods” allowed them to conduct research that they consider “empiricallyvalid and historically meaningful. More than ever, we believe that there isindeed no methodological justification to limit ourselves to a single, ‘fixedmethodology’” (p. 24).

Several advantages exist for combining the quantitative “cold statistics”and the qualitative “richly detailed stories.” Such a combination has the

DESIGNING COMMUNITY-BASED MIXED METHODS RESEARCH 43


Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

advantages of being favorable to community members, allowing the communityto have greater control and power in the research process, reducing surveyerror, providing stronger evidence, addressing weaknesses of any one approach,and disseminating the findings in a more accessible way (Axinn & Pearce,2006; Jacklin & Kinoshameg, 2008; Yardley & Bishop, 2008). With fundingagencies increasingly emphasizing mixed methods research, the use of com-bined qualitative and quantitative approaches enlarges the amount of evi-dence available to understand complex research problems (Creswell & PlanoClark, 2011). Thus, building on the current practices of gathering both formsof data in a community-based study and taking advantage of the strengths ofeach naturally leads to the use of mixed methods research.

MIXED METHODS RESEARCH STUDIES

Mixed methods research involves the collection, analysis, and mixing ofboth qualitative and quantitative data in a single study or a multiphase programof inquiry (see Creswell & Plano Clark, 2011). It includes persuasive and rig-orous qualitative and quantitative procedures, the use of worldviews to providea foundation for the research, and the application of specific types of designs.Different forms of relating the qualitative and quantitative data can be shapedinto specific research designs. Many types of designs have been advanced in themixed methods literature since the late 1980s (e.g., Greene, Caracelli, &Graham, 1989; Morse, 1991; Tashakkori & Teddlie, 1998). Creswell and PlanoClark’s (2011) classification system includes six major types of designs andprovides a parsimonious, workable set of designs currently used in the mixedmethods field. Although each of these is briefly introduced here, readers areencouraged to refer to Creswell and Plano Clark (2011) for an in-depth under-standing of mixed methods research.

A key to understanding these designs is to consider whether the inves-tigator collects the quantitative and qualitative data concurrently (at roughlythe same time) or sequentially (with one database following another). A pop-ular concurrent design is the convergence parallel design, which involves col-lecting both quantitative and qualitative data separately and independentlyand then merging the two forms of data at the same time to understand aresearch problem or question. The explanatory sequential design involvescollecting data sequentially with the first phase as a quantitative phase andthe second phase as a qualitative phase. The intent of this design is that thesecond phase qualitative data helps to explain the first phase quantitativeresults. The exploratory sequential design also involves collecting data sequen-tially but in an order reversed from the explanatory sequential design. Inthe exploratory design, the investigator begins by collecting and analyzing



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

qualitative data, and this is followed by a second phase of collecting and ana-lyzing quantitative data. The key idea here is that an initial qualitative explo-ration can help to best develop a quantitative assessment that might begeneralized to a large sample of a population. The embedded design is anotherdesign used by the mixed methods researcher in collecting and analyzing bothconcurrent and sequential data. In this design, a secondary database providessupport for a primary database to develop a more complete understanding orto facilitate the use of the primary database. A transformative design can alsoemploy quantitative and qualitative data concurrently or sequentially. In thiscase, a transformative perspective surrounds all phases of the mixed methodsdesign and provides a lens for the study as well as a framework for providingsuggestions for change. A final type of mixed methods design is the multi-phase design, which involves multiple phases or projects conducted overtime with both concurrent and sequential data collection and analysis pro-cedures. An example of this design would be the multiproject evaluationstudies in which both quantitative and qualitative data are collected and ana-lyzed through the stages of a needs assessment, the development of an inter-vention program, and the assessment of the program. These six designs thusbecame the mixed methods approaches that were salient in our examinationof community-based studies.

A REVIEW OF MIXED METHODS COMPONENTS INCOMMUNITY-BASED RESEARCH STUDIES

We conducted a literature review to identify community-based studiesthat involved the collection and analysis of both forms of data. Using the searchterms mixed methods, community-based research, ethnocultural, and methodology,we sought to identify studies that were community-based and that includedquantitative and qualitative data collection and analysis. However, becausethe term ethnocultural did not yield any studies that were both community-based or mixed methods research, we expanded our inclusion criterion tolarger populations, taking into consideration ethnically diverse communitiesat the lower end of the socioeconomic scale.

Incorporation of Qualitative and Quantitative Data

As shown in Table 3.1, we found 10 community-based studies thatcollected qualitative and quantitative data. In these studies, the type of datacollected was overwhelmingly qualitative. Of 46 instances of data collection,36 were qualitative. These researchers used a wide variety of strategies. Themost common qualitative type of data collection were interviews (12 of 36),



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.


TA

BLE

3.1

Mix

ed M

etho

ds C

hara

cter

istic

s of

the

Com

mun

ity-B

ased

Mix

ed M

etho

ds R

esea

rch

Art

icle

s

Sou

rce

Com

mun

ityQ

ualit

ativ

eQ

uant

itativ

eT

imin

gS

tage

Aro

nson

et a

l., 2

007

Jack

lin &

Kin

osha

meg

,20

08

Mos

avel

et a

l., 2

005

Nay

lor e

t al.,

200

2

Rad

da e

t al.,

200

3

J. J

. Sch

ensu

let

al.,

200

6

S. L

. Sch

ensu

l, N

asta

si,

& V

erm

a, 2

006

Teu

fel-S

hone

et a

l., 2

006

Wes

thue

s et

al.,

200

8

Woo

ds, 2

009

Rec

ipie

nts

of

com

mun

ity-b

ased

prog

ram

Wik

wem

ikon

g In

dian

s

Sou

th A

fric

ans

Rec

ipie

nts

of

com

mun

ity-b

ased

heal

th in

terv

entio

nO

lder

urb

an a

dults

in h

igh-

HIV

-ris

kne

ighb

orho

ods

Low

-inco

me

Indi

anm

en

Low

-inco

me

Indi

anm

en

Hua

lapa

i

Fiv

e et

hnic

gro

ups

Bla

ck A

mer

ican

s

Foc

us g

roup

s, k

ey

info

rman

t int

ervi

ews,

neig

hbor

hood

map

ping

,jo

urna

lsE

xten

sive

par

ticip

ant

obse

rvat

ion

Foc

us g

roup

s, in

form

alin

terv

iew

sF

ocus

gro

ups

Eth

nogr

aphi

c do

cum

enta

-tio

n of

bui

ldin

gs a

ndne

ighb

orho

ods,

obs

erva

-tio

ns, i

nfor

mal

inte

rvie

ws,

sem

istru

ctur

ed in

terv

iew

sE

licita

tion

tech

niqu

es, o

ral

hist

ory,

eth

nogr

aphi

cin

terv

iew

s an

d ob

serv

a-tio

ns, s

emis

truc

ture

d, in

-de

pth

inte

rvie

ws

Key

info

rman

t int

ervi

ews,

in-d

epth

inte

rvie

ws,

soci

al a

nd g

eogr

aphi

cm

appi

ngS

emis

truct

ured

inte

rvie

ws,

self-

asse

ssm

ent t

ool

Sem

istr

uctu

red

inte

rvie

ws,

focu

s gr

oups

Key

info

rman

t int

ervi

ews

and

focu

s gr

oups

Com

mun

ity s

urve

ys

Com

mun

ity n

eeds

asse

ssm

ent

surv

eyS

urve

ys

Wor

kshe

et

Epi

dem

iolo

gica

lsu

rvey

Sta

ndar

dize

dsc

reen

ing

tool

s

Sur

vey

Loca

lly g

ener

ated

envi

ronm

enta

lin

vent

ory

Sur

vey

Sur

vey

Seq

uent

ial

Con

curr

ent

Con

curr

ent

Con

curr

ent

Con

curr

ent

Seq

uent

ial

Seq

uent

ial

Con

curr

ent

Con

curr

ent

Con

curr

ent

Eva

luat

ing

Con

nect

ing,

diag

nosi

ng

Con

nect

ing,

diag

nosi

ngE

valu

atin

g

Dia

gnos

ing

Dia

gnos

ing,

eval

uatin

g

Dia

gnos

ing

Dia

gnos

ing

Dia

gnos

ing

Pre

scrib

ing/

impl

emen

ting


Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

followed by focus groups (five of 36), and observations (three of 36). Researchersdescribed their interviews as “ethnographic” (J. J. Schensul et al., 2006), “in-depth” (J. J. Schensul et al., 2006; S. L. Schensul et al. 2006), “informal”(Mosavel et al., 2005; Radda et al., 2003), and “semistructured” (Radda et al.,2003; J. J. Schensul et al., 2006; Teufel-Shone et al., 2006; Westhues et al.,2008). Several researchers mentioned that their interviews were with “keyinformants” (Aronson et al., 2007, p. 103; S. L. Schensul et al., 2006, p. 102;Teufel-Shone et al., 2006, p. 1623; Woods, 2009, p. 258). Several researcherscollected qualitative data about community neighborhoods. Radda et al.(2003) used ethnographic documentation of buildings and neighborhoods.Aronson et al. (2007) engaged in neighborhood mapping, whereas S. L.Schensul et al. (2006) used social and geographic mappings of communities.There were several other types of qualitative data collected, such as elicita-tion techniques (J. J. Schensul et al., 2006), journals (Aronson et al., 2007),a self-assessment tool that involved group discussions (Teufel-Shone et al.,2006), and oral history (J. J. Schensul et al., 2006).

The remaining instances of data collection were quantitative (10 of 46).The most common type of data was surveys (seven of 10; for an example, seeWesthues et al., 2008). Surveys were further described as community (Aronsonet al., 2007) or epidemiological (Radda et al., 2003). The community needsassessment survey used by Jacklin and Kinoshameg (2008) contained open-ended items in addition to quantitative items. Other forms of quantitative datainclude a locally generated environmental inventory (Teufel-Shone et al.,2006), standardized screening tools (J. J. Schensul et al., 2006), and a work-sheet (Naylor et al., 2002). Thus, the trend we identify in community-basedstudies is collecting primarily qualitative data, and the studies we found onlycollected one form of quantitative data per project.

Mixed Methods Components Incorporated Into Community-Based Studies

We analyzed the empirical articles according to mixed methods criteriaset forth by Creswell and Plano Clark (2011): whether the researcher specif-ically identified a mixed methods design, whether they provided a rationalefor collecting multiple forms of data, and whether they presented detailedprocedures in their methodology sections.

Creswell and Plano Clark (2011) recommended describing a study inmixed methods terms and specifying the type of mixed methods design used.In the 10 studies we found, only three used the term mixed methods (see, e.g.,S. L. Schensul et al., 2006). Additionally, only two articles addressed the typeof mixed methods design. Woods (2009) used an exploratory sequentialdesign (see Creswell & Plano Clark, 2011) and Westhues and colleagues(2008) described their design as a “collaborative mixed method participatory



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

action research design” (p. 702) that was exploratory and qualitatively driven.However, none of the researchers provided detail on the design or their reasonsfor using it.

We also examined the studies to determine whether the authors provideda rationale for collecting both qualitative and quantitative data. Althoughcommunity-based researchers typically explained why they collected individualforms of data, most did not explain why they collected multiple types of data.Even when the researchers did mention their rationale for mixing, they did notprovide much detail. For example, Naylor et al. (2002) indicated that theyexpanded their quantitative study to include qualitative approaches to improve“community receptivity” but did not provide further information (p. 1175).The emergent nature of data collection was also indicated by some but notother researchers. For instance, Woods (2009) explained that the communitydetermined what forms of data to collect. Similarly, J. J. Schensul and colleagues(2006) described how their lead agency of the study contributed to the mixedmethods experience.

Finally, Creswell and Plano Clark (2011) emphasized that researchersshould report detailed procedures and suggested creating a visual diagramof procedures. Few community-based studies reported detailed procedures.However, there were exceptions. Aronson et al. (2007) included tables oftheir data collection and the purposes for collecting each form of data.Westhues et al. (2008) provided detailed descriptions of their data collectionprocedures. Although they did not provide a diagram of their procedures, theydid include a table of the conceptual framework for their focus groups.

Although rigor is advanced as a core element of mixed methods research(Creswell & Plano Clark, 2011) based on design, the rationale for collectingmultiple forms of data, and the detailed procedures, rigor was not apparentin the community-based studies we reviewed. This led us to inquire furtheras to the use of qualitative and quantitative data in community-based stud-ies and to make suggestions as to how they might be framed within mixedmethods designs.

Framing Mixed Methods Designs Within a Community-BasedTheoretical Model

One community-based theoretical model seemed attractive for exam-ining the integration of multiple forms of data and mixed methods designs.The model by Stoecker (2005) referred to collecting multiple forms of data,applying the data across disciplines, and encompassing stages that comple-mented the empirical studies we had identified in the literature. Thus, we pro-pose an adaptation of Stoecker’s model for conducting CBMMR, as shown inFigure 3.1. We made two modifications in Stoecker’s model to better reflect



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

the community-based studies we had seen in the literature. Stoecker’s originalmodel included four stages: diagnosing, prescribing, implementing, andevaluating. However, we decided to combine prescribing and implement-ing into one stage and to add two stages that highlight starting and endinga CBMMR project: connecting and disseminating. Thus, in building onStoecker’s work, the stages that we describe are the following: connecting,diagnosing, prescribing–implementing, evaluating, and disseminating.

Connecting Stage

Stoecker (2005) stated that one of the first steps a community organizermust take is getting to know the community. He proposed diagnostic researchsuch as knocking on residents’ doors to ask what concerns them and identi-fying the community’s most concerned members. In practice, community-based researchers have expanded on this initial step to describe a lengthyprocess of developing lasting relationships built on trust and ongoing commu-nication before diagnosing. In Figure 3.1, this stage is exemplified by a circle to


Figure 3.1. Recommended working model for community-based mixed methodsresearch.

Prescribing/implementing Recommended data: Interviews and surveys Recommended design: Exploratory sequential m odel

Evaluating Recommended data: Focus groups and pre- and postsurveys Recommended design: Em bedded experi me ntal model

Connecting Recommended data: Unobtrusive observations and informal interviews Recommended design: Instrument developm ent m odel

Communi t y

Disseminating Recommended data: Interviews and surveys Recommended design: Exploratory taxono my developm ent m odel

Social, political, and econom ic contexts

Diagnosing Recommended data: Focus groups, neighborhood ma pping, and surveys Recommended design: Triangulation design: Convergence model


Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

indicate that it is a dynamic, constant process. Thus, the connecting stage isa relationship-building stage characterized by interactions with community.

Because it is easy for researchers to misinterpret the community’s inten-tions, concerns, language, or other ethnocultural aspects, we recommend thatresearchers consistently use strategies in this stage to understand their biases. Itis likely that formal research cannot be done immediately. However, mixedmethods researchers can use other methods by checking with community mem-bers and stakeholders, keeping journals, and bracketing as Mosavel et al. (2005)did in their study on an underprivileged community in Cape Town, SouthAfrica. Being flexible and patient is crucial in connecting. At this stage, we rec-ommend unobtrusive observations. Jacklin and Kinoshameg (2008) used thisapproach in their study with Wikwemikong Indians in Ontario, Canada. Infor-mal interviewing can also be used (see Mosavel et al., 2005). The connectingphase is especially important in communities that are mistrustful of researchor universities (Jacklin & Kinoshameg, 2008; Mosavel et al., 2005). If theresearcher does use quantitative methods, they should be informed by the qual-itative findings. One potential design is the exploratory sequential design withthe intent of developing an instrument (Creswell & Plano Clark, 2011). In thefirst phase the researcher qualitatively explores the topic and in the secondphase uses the qualitative findings to develop items for a quantitative surveyand then administers the survey. Although we did not find an empirical exam-ple of this design, we believe it would be a good fit for this stage.

Diagnosing Stage

Diagnosing involves exploring the community’s concerns to understandtheir key problems and to identify opportunities for change (Stoecker, 2005).The purpose of diagnosing is twofold: to assess individuals and their environ-ments. Particularly for ethnocultural communities, understanding the contextof the participants’ experiences is important. Qualitative approaches to datacollection can include focus groups, which help to provide understanding of thecommunity’s perspectives in a social setting. Focus groups may also feel morecomfortable as an initial form of data collection, as in Westhues et al.’s (2008)study with five ethnocultural groups in Ontario, Canada. They selected focusgroups because they contribute to “multimethod, multiperspective theorybuilding . . . while engaging multicultural community participants in an inter-active data collection method” (p. 707). Westhues et al.’s study of five ethno-cultural groups provide an example of this diagnosing using mixed methodsresearch. They conducted focus groups to explore viewpoints on mental healthproblems, interventions, and services that were needed. They also administereda survey that included open-ended items to study the organizations’ perceptionsof how well they were responding to the communities’ needs. They used theirquantitative analysis of the survey to confirm their focus group findings.



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

With regard to the mixed methods design that is most suitable for diag-nosing, an embedded design is useful because researchers can remain open tothe needs of the community while supplementing their conclusions with quan-titative data. By embedding quantitative data into qualitative data, it seemsthat a more comprehensive diagnosis of the community’s issue can emerge.On the other hand, by building on qualitative research to develop quantitativesurveys, Teufel-Shone et al.’s (2006) study demonstrates how the sequentialdesign can be used when a culturally sensitive assessment is needed. Theirproject is an example of developing survey questions that emerge from the com-munity. Even more specifically, they used locally generated surveys to under-stand the Hualapai community in northwest Arizona. Developing culturallysensitive measures is crucial for ethnocultural groups that have traditionallybeen marginalized.

Prescribing and Implementing Stages

The second and third stages of Stoecker’s (2005) model are prescribingand implementing, and they are discussed together because of their intersect-ing nature. The prescribing stage refers to the research planning that is requiredbefore conducting the study and finalizing the study design. The next stage isimplementing, and the research itself becomes the project, serving as boththe process and outcome goal. Because few examples were found in the extantliterature on prescribing and implementing, only two empirical examples arepresented, both of which are exemplars of how the prescribing and imple-menting process can look in a community-based mixed methods study.

Prescribing, known as the research planning process, has to do with thepreparations that are necessary before carrying out a research study. Prescrib-ing can also be described as planning and is the effort of matching the com-munity needs with practical considerations such as whether the research goalis directly relevant and compatible with the community. Information gather-ing plays a critical role in this stage and includes literature reviews as well asgathering different sources of knowledge about the study content. In the pre-scribing process, the goal for making social change research must also be clar-ified, either as internally (i.e., community-betterment) or externally focused(i.e., advancement of existing policies).

The prescribing/implementing stage involves doing the operationalresearch on an intervention, and then developing the intervention and put-ting it into practice. Although the need for a thorough literature review is anecessity in every research endeavor, conducting CBMMR requires an evenmore comprehensive examination and review of (a) the needs of the commu-nity (which is separate from the research on intervention), (b) the topic ofstudy as it relates to other communities with similar ethnographic profiles andneeds, and (c) the applicability and interaction of both. A thorough literature



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

review may not always be feasible because of limitations such as communityconsiderations, funding, and project timeline. One possible design is anexploratory sequential design. Researchers can apply this approach by usingthe initial quantitative data collection to inform the second, qualitativephase.

As for the implementing stage, Woods’s (2009) study about engagingAfrican American communities in health planning serves as an ideal exam-ple of research being the action in the study. The research project becomesthe intervention in asking the question, “How can we get African Americansinvolved in research and prevention activities” (p. 247)? Furthermore, the authorexplicitly identifies the purpose and process of this project in the abstract, such that“The African American Health Initiative represents an answer to this question andis in many ways a success story of a local Black population that willingly engagedin health planning through a community-based participatory research process. As asuccess story, the research project served to advocate for the engagement of AfricanAmerican communities to be involved and collaborative in health planning” (p. 247).

In Woods’s (2009) study, multiple perspectives and methods of data col-lection were used, with the voices serving to empower the community andwith the involvement itself being the act of empowerment. Although notstated by the researcher, this project fits within the framework of the trans-formative design, in which the research itself is aimed to make significantsocial changes for marginalized populations. In addition to the long-lastingchanges of the research even after the study was completed, it is clear thatWoods’s study is guided by the purpose of highlighting the marginalization ofAfrican Americans in health planning and that this theme is central acrossall of the phases of the research project, from the introduction to the end.Even in the conclusion of the study, Woods advocated for addressing theneeds of African Americans, underscoring the limitations of the health caresystem for this particular group of individuals.

In the Woods (2009) study, it is evident that both the quantitative andqualitative methods used in this study were able to contribute to advancingthe needs of a specific community while also generalizing to larger communi-ties. The implications of this study underscore the uniqueness of mixed meth-ods research and, more specifically, CBMMR because of the ways in whichthe quantitative and qualitative findings were able to advance one commu-nity. If the study had been purely qualitative or quantitative, the findingswould not have attended to the specific needs of one particular AfricanAmerican community and give voice to the people who have not been heardor asked about their health planning. At the same time, however, it is becauseof the quantitative data in this study that there is greater understanding of thehealth issues and barriers for the African American community and the pos-



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

sibility of generalizing the findings from this study to other marginalizedgroups and communities. Altogether, integrating both quantitative and qualita-tive data in this study demonstrated the powerful nature of gathering multipleperspectives through multiple sources in one phase to empower a communitythrough the research process.

Evaluating Stage

Evaluating a study consists of assessing the intervention’s impact on thecommunity. Stoecker (2005) stated that the purpose of evaluating an inter-vention is to improve implementation, and thus the stages in Figure 3.1 arelinked by double arrows. Focus groups were recommended at the evaluatingstage (for an example, see Naylor et al., 2002). This type of data collectionprovided a way for the community to discuss their shared experiences of theintervention. One example was Aronson and colleagues’ (2007) use of focusgroups to identify contextual variables and neighborhood features and exploreperceptions of local concerns, strengths, and salient beliefs of recipients ofcommunity-based programs in underserved areas of Baltimore. Quantitativedata recommended at this stage were surveys that indicate how participantshave changed. Naylor et al.’s (2002) worksheet examined participants’ per-spectives on a health intervention in British Columbia, Canada. Especiallyhelpful are pre- and postintervention surveys. Thus, one design that fits thisstage is the embedded design within an experimental framework (Creswell &Plano Clark, 2011). Basically, the data process would consist of qualitativedata in a supporting role in an experimental design. However, CBMMR studieswith strong quantitative components should be conducted with cautionbecause they can potentially result in the community’s inability to determineits own realities (Jacklin & Kinoshameg, 2008). Thus, the embedded designwithin an experimental framework should be used only when trust has beengenerated within the community and depth of understanding about its membershave been gained.

Disseminating Stage

Community partners have expressed concern over lack of use ofresearch results (Jacklin & Kinoshameg, 2008). For this reason, disseminat-ing is essential to community-based research. Disseminating the findingsincludes assessing the influence of the study on the community members. Inour model, disseminating might occur at two points: at the diagnosing stageand with the community at the end of the project. The process should be cir-cular and iterative and should respond to the needs of the community. Sec-ond, it should produce social change at the community level and within thesociopolitical contexts in which community members reside. An example of



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

disseminating is in Woods’s (2009) study on African American health plan-ning. Following the completion of the research project, the findings continuedto affect the community through the formation of committees and nonprofitorganizations aimed to create self-help programs communities. Furthermore,the knowledge gained with this community was even adapted by anotherethnic minority group—thus influencing multiple communities beyond theone that was studied.

Qualitative interviews could explore what elements of disseminatingthe findings were most beneficial to community members. A follow-up quan-titative survey could investigate the generalizability of those helpful elements.Thus, an exploratory design with the intent of developing a taxonomy can beused by qualitatively exploring a topic followed by disseminating it in a waythat is effective to participants. These findings can then guide the develop-ment of items for quantitative survey.

Altogether, framing a project as mixed methods would enhance the use ofrigorous procedures in community-based research. It could also provide modelsfor conceptualizing studies for members of a research team, community membersinvolved in the project, and policy makers who need to support the project.

RECOMMENDATIONS

On the basis of our analysis, we offer the following recommendations forresearchers interested in conducting CBMMR.

Specify the Research Design

We recommend that researchers in community-based projects describetheir methods in detail. Although researchers collected both forms of data, theygenerally elaborated on only one form in their results section (Radda et al., 2003;Westhues et al., 2008), with some exceptions (e.g., Woods, 2009). Also, it wouldbenefit readers to cite a mixed methods design, such as Woods’s work (2009),which was self-identified as a “mixed-method exploratory sequential” study(p. 254). Such an explicit description of the research design not only helps toinform the reader of the procedures of the study but also allows for study replica-tion to encourage researchers to build on the current study in other commu-nities or content areas. Furthermore, it may help community-based researchersbe more mindful of the mixed methods aspects of their projects.

Explain the Value of Each Data Source

Because the results from community-based research studies have impli-cations for policies, practices, and communities, we suggest that researchers



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

articulate how each form of data they collect contributes to the study findings.Westhues et al.’s (2008) study serves as an example because they describedwhat their study gained from the qualitative and quantitative findings. Theysubsequently highlighted the advantages of collecting both qualitative andquantitative data through triangulating different sources of data, having onesource complement the other, and gaining new insights through a dialecticapproach. In some of the studies in which recommendations were presented,the suggestions advanced by the researchers seemed to be driven only by thequalitative findings (e.g., Radda et al., 2003). Generalizing the findings toother communities may therefore be inappropriate given that the qualitativeresearch provides depth, whereas quantitative research can generalize thestudy findings. If authors are more intentional about their designs, their useof mixed methods will carry more weight and can be replicated and advancedin theory and practice.

Explain the Purpose of Using Multiple Forms of Data

Consistent with our previous recommendation, we urge researchers whoare conducting CBMMR to specify their reasons for collecting multiple formsof data and to explicitly explain their rationale in each step of their researchprocess, beginning at the research design stage to the discussion section. By beingmore explicit about the research design, readers can provide a better understand-ing of how the different forms of data serve to answer the research question(s)(see S. L. Schensul et al., 2006) as well as the benefits of using mixed methodsapproaches in a community-based study (Jacklin & Kinoshameg, 2008). In addi-tion, justifying the use of a mixed methods approach in the discussion sectioncan underscore the value of using multiple methods to arrive at multiple per-spectives and multidimensional findings.

Use Different Mixed Methods Research Designs

We believe that there is a wide array of research designs that can be used.In particular, the multiphase design seems to be underused in community-based research and yet ideal for CBMMR because of its longitudinal nature aswell as the ability to combine concurrent and sequential data collection. Con-ducting a community-based study using a multiphase design can encompassseveral stages of community-based research, such as connecting, diagnosing,and prescribing/implementing. Several community-based studies did, in effect,implement a multiphase design (see Jacklin & Kinoshameg, 2008; Mosavelet al., 2005). We believe that a multiphase design would allow for more breadthas well as depth of the study content and allow expand on the current researchdesigns to best address the needs of the ethnocultural community.



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

Integrate Mixed Methods With Community-Based Research

Finally, with regard to our recommendations for conducting CBMMR,we advocate for more discussions and dialogues regarding the use of mixedmethods in community-based research and of the applicability of conductingcommunity-based projects in mixed methods research. In our process of iden-tifying studies that explicitly addressed and met the criteria of being both acommunity-based study and a mixed methods research study, we found thatthere was minimal overlap in the existing empirical database. Rarely did wesee the term mixed methods in any of the literature on community-basedresearch, although the issue of methodology, as a whole, was popularly dis-cussed as it relates to community-based research. We understand that in con-ducting community-based research, the logistical challenges along withethical dilemmas faced by researchers and community members can be daunt-ing and overwhelming; thus, writing about the methodologies may seemexcessive. At the same time, however, we underscore the value of elaborat-ing on the research design and rationale of a study, given its direct impact onthe research question and the goals of the study.

An argument against the present approach is that community-basedresearchers cannot emphasize methodology above the rights of the commu-nity. We wholeheartedly agree that the rights of the community membersmust be respected at all times. However, we believe that, when possible, rig-orous data collection, analysis, and mixing should be employed. There areseveral reasons for this position. First, detailing methods can provide guid-ance to others who want to work within that community or similar commu-nities. Second, making the most of mixed methods designs ensures thatresearchers are using the information they glean to the fullest. The way thatresearch is conducted is an important element, and so are the participants’responses to the different methods. Third, the use of rigorous mixed methodsmay be favorable if the research is related to policy (Jacklin & Kinoshameg,2008). Furthermore, we argue that both the research and the interventionshould be consistently reevaluated.

The aim of this chapter was to offer a unique contribution to the liter-ature by highlighting approaches to incorporating mixed methods researchinto community-based studies. Our chapter combined mixed methods withcommunity-based approaches to provide theoretical and practical guidance.Theoretically, our CBMMR model provided a conceptual framework forunderstanding CBMMR. Practically, we guided researchers by highlightingspecific data collection and mixed methods strategies and designs. By adapt-ing an existing community-based model to highlight the importance of mixedmethods in community-based research, we sought to provide a systematicapproach to conducting CBMMR. It is hoped that the strategies provided in



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

the model can serve as tools for community-based researchers at all experiencelevels. Although much has been done in the existing empirical studies thatwe reviewed, we used our analysis of the current empirical studies to offer sug-gestions for ways that community-based researchers can employ mixed methodsinto their studies. We believe that CBMMR can be a valuable methodologi-cal and research approach to address social concerns and advance a uniqueconversation about the intersection of mixed methods and community-basedresearch.

REFERENCES

Andrew, S., & Halcomb, E. J. (2006). Mixed methods research is an effective methodof enquiry for community health research. Contemporary Nurse, 23, 145–153.doi:10.5172/conu.2006.23.2.145

Aronson, R. E., Wallis, A. B., O’Campo, P. J., Whitehead, T. L., & Schafer, P.(2007). Ethnographically informed community evaluation: A framework andapproach for evaluating community-based initiatives. Maternal and Child HealthJournal, 11, 97–109. doi:10.1007/s10995-006-0153-4

Axinn, W. G., & Pearce, L. D. (2006). Mixed method data collection strategies. NewYork, NY: Cambridge University Press. doi:10.1017/CBO9780511617898

Boynton, P. M. (2002). Life on the streets: The experiences of community researchersin a study of prostitution. Journal of Community & Applied Social Psychology, 12,1–12. doi:10.1002/casp.657

Creswell, J. W., & Plano Clark, V. (2011). Designing and conducting mixed methodsresearch (2nd ed.). Thousand Oaks, CA: Sage.

Greene, J. C., Caracelli, V. J., & Graham, W. F. (1989). Data analysis strategies formixed-method evaluation designs. Educational Evaluation and Policy Analysis,15, 195–207.

Harris, G. E. (2006). Practicing HIV/AIDS community-based research. AIDS Care,18, 731–738. doi:10.1080/09540120500307735

Israel, B. A., Schulz, A. J., Parker, E. A., & Becker, A. B. (1998). Review of community-based research: Assessing partnership approaches to improve public health.Annual Review of Public Health, 19, 173–202. doi:10.1146/annurev.publhealth.19.1.173

Jacklin, K., & Kinoshameg, P. (2008). Developing a participatory Aboriginal healthresearch project: “Only if it’s going to mean something.” Journal of EmpiricalResearch on Human Research Ethics, 3, 53–67. doi:10.1525/jer.2008.3.2.53

Keys, C. B., McMahon, S., Sanchez, B., London, L., & Abdul-Adil, J. (2004). Cul-turally anchored research: Quandaries, guidelines, and exemplars for commu-nity psychology. In L. A. Jason, C. B. Keys, Y. Suarez-Balcazar, R. R. Taylor, &M. I. Davis (Eds.), Participatory community research: Theories and methods in



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

action (pp. 177–198). Washington, DC: American Psychological Association.doi:10.1037/10726-010

Kidder, L. H., & Fine, M. (1987). Qualitative and quantitative methods: When storiesconverge. In M. M. Mark & R. L. Shotland (Eds.), Multiple methods in programevaluation: New directions in program evaluation (pp. 57–75). San Francisco, CA:Jossey-Bass.

Linney, J. A. (2000). Assessing ecological constructs and community context. In J. Rappaport & E. Seidman (Eds.), Handbook of community psychology(pp. 647–668). New York, NY: Plenum Press. doi:10.1007/978-1-4615-4193-6_27

Morse, J. M. (1991). Approaches to qualitative-quantitative methods: Applicationsto health research. Qualitative Health Research, 8, 362–376.

Mosavel, M., Simon, C., van Stade, D., & Buchbinder, M. (2005). Community-basedparticipatory research (CBPR) in South Africa: Engaging multiple constituentsto shape the research question. Social Science & Medicine, 61, 2577–2587.doi:10.1016/j.socscimed.2005.04.041

Naylor, P. J., Wharf-Higgins, J., Blair, L., Green, L., & O’Connor, B. (2002). Evalu-ating the participatory process in a community-based heart health project. SocialScience & Medicine, 55, 1173–1187. doi:10.1016/S0277-9536(01)00247-7

Radda, K. E., Schensul, J. J., Disch, W. B., Levy, J. A., & Reyes, C. Y. (2003). Assessinghuman immunodeficiency virus (HIV) risk among older urban adults: A model forcommunity-based research partnership. Family & Community Health, 26, 203–213.

Robins, C. S., Ware, N. C., dosReis, S., Willging, C. E., Chung, J. Y., & Lewis-Fernández, R. (2008). Dialogues on mixed-methods and mental health servicesresearch: Anticipating challenges, building solutions. Psychiatric Services, 59,727–731. doi:10.1176/appi.ps.59.7.727

Sasao, T., & Sue, S. (1993). Toward a culturally anchored ecological framework ofresearch in ethnic-cultural communities. American Journal of Community Psy-chology, 21, 705–727. doi:10.1007/BF00942244

Schensul, J. J., Robison, J., Reyes, C., Radda, K., Gaztambide, S., & Disch, W. (2006).Building interdisciplinary/intersectoral research partnerships for community-based mental health research with older minority adults. American Journal ofCommunity Psychology, 38, 79–93. doi:10.1007/s10464-006-9059-y

Schensul, S. L., Nastasi, B. K., & Verma, R. K. (2006). Community-based researchin India: A case example of international and transdisciplinary collaboration.American Journal of Community Psychology, 38, 95–111. doi:10.1007/s10464-006-9066-z

Sirin, S. R., Bikmen, N., Mir, M., Fine, M., Zaal, M., & Katsiaficas, D. (2008). Explor-ing dual identification among Muslim-American emerging adults: A mixedmethods study. Journal of Adolescence, 31, 259–279. doi:10.1016/j.adolescence.2007.10.009

Sirin, S. R., & Fine, M. (2007). Hyphenated selves: Muslim American youth nego-tiating identities on the fault lines of global conflict. Applied Developmental Sci-ence, 11(3), 151–163. doi:10.1080/10888690701454658



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

Sirin, S., & Fine, M. (2008). Muslim American youth: Understanding hyphenated iden-tities through multiple methods. New York, NY: New York University Press.

Sterk, C. E., & Elifson, K. W. (2004). Qualitative methods in community-basedresearch. In D. S. Blumenthal & R. J. DiClemente (Eds.), Community-basedhealth research (pp. 133–152). New York, NY: Springer.

Stoecker, R. (2005). Research methods for community change: A project-based approach.Thousand Oaks, CA: Sage.

Strand, K., Marullo, S., Cutforth, N., Stoecker, R., & Donohue, P. (2003). Community-based research in higher education: Principles and practices. San Francisco, CA:Jossey-Bass.

Sue, D. W., & Sue, D. (2003). Counseling the culturally diverse: Theory and practice(4th ed.). New York, NY: Wiley.

Tashakkori, A., & Teddlie, C. (1998). Mixed methodology: Combining qualitative andquantitative approaches. Thousand Oaks, CA: Sage.

Taylor, R. R., Jason, L. A., Keys, C. B., Suarez-Balcazar, Y., Davis, M. I., Durlak, J. A.,& Isenberg, D. H. (2004). Introduction: Capturing theory and methodology in par-ticipatory research. In L. A. Jason, C. B. Keys, Y. Suarez-Balcazar, R. R. Taylor, &M. I. Davis (Eds.), Participatory community research: Theories and methods in action(pp. 177–198). Washington, DC: American Psychological Association.

Teufel-Shone, N. I., Siyuja, T., Watahomigie, H. J., & Irwin, S. (2006). Community-based participatory research: Conducting a formative assessment of factors thatinfluence youth wellness in the Hualapai community. American Journal of Pub-lic Health, 96, 1623–1628. doi:10.2105/AJPH.2004.054254

Trimble, J. E., & Fisher, C. (2006). The handbook of ethical research with ethnoculturalpopulations & communities. Thousand Oaks, CA: Sage.

Westhues, A., Ochocka, J., Jacobson, N., Simich, L., Maiter, S., Janzen, R., & Fleras,A. (2008). Developing theory from complexity: Reflections on a collaborativemixed method participatory action research study. Qualitative Health Research,18, 701–717. doi:10.1177/1049732308316531

Woods, V. D. (2009). African American Health Initiative Planning Project: A socialecological approach utilizing community-based participatory research methods.Journal of Black Psychology, 35, 247–270. doi:10.1177/0095798409333589

Yardley, L., & Bishop, F. (2008). Mixing qualitative and quantitative methods: Apragmatic approach. In C. Willig & W. Stainton-Rogers (Eds.), The Sage hand-book of qualitative research in psychology (pp. 352–369). Thousand Oaks, CA:Sage. doi:10.4135/9781848607927.n20

Zeller, R. A. (1993). Combining qualitative and quantitative techniques to developculturally sensitive measures. In D. G. Ostrow & R. C. Kessler (Eds.), Method-ological issues in AIDS behavioral research (pp. 95–116). New York: Plenum Press.



Copy

righ

t Am

eric

an P

sych

olog

ical

Ass

ocia

tion

. No

t fo

r fu

rthe

r di

stri

buti

on.

Documents

DESIGNING COMMUNITY-BASED MIXED METHODS RESEARCHsites.bu.edu/miccr/files/2017/03/Badiee-2012-Designing-Community... · designing community-based mixed methods research ... conducting