Received 07/31/11

Revised 04/13/12
Accepted 09/04/12
DOI: 10.1002/j.1556-6676.2013.00085.x

Administering Quantitative Instruments

With Qualitative Interviews:
A Mixed Research Approach
Rebecca K. Frels and Anthony J. Onwuegbuzie

The authors demonstrate how collecting quantitative data via psychometrically sound quantitative instruments dur-
ing the qualitative interview process enhances interpretations by helping researchers better contextualize qualitative
findings, specifically through qualitative dominant crossover mixed analyses. They provide an example of this strategy,
whereby a baseline was established using a quantitative scale and normative data to help interpret qualitative inter-
views, resulting in what they call a mixed methods interview. Philosophical and practical implications are discussed.

Keywords: qualitative interviews, qualitative dominant mixed analysis, crossover mixed analysis, mixed methods
research, mixed research

For practitioners in the field of counseling, the combining or It recognizes the importance of traditional quantitative and
mixing of qualitative and quantitative methodologies is not a qualitative research but also offers a powerful third paradigm
new or unique phenomenon. In fact, as surmised by Powell, choice that often will provide the most informative, complete,
Mihalas, Onwuegbuzie, Suldo, and Daley (2008), “by defini- balanced, and useful research results. Mixed methods research
tion, assessment, whether for purposes of program planning or is the research paradigm that (a) partners with the philosophy
treatment, necessitates the consideration of multiple sources of of pragmatism in one of its forms (left, right, middle); (b)
data” (p. 293). In addition, as stated in Section E (i.e., Evalua- follows the logic of mixed methods research (including the
tion, Assessment, and Interpretation) of the ACA Code of Ethics logic of the fundamental principle and any other useful logics
(American Counseling Association, 2005), counselors use both imported from qualitative or quantitative research that are
quantitative and qualitative assessments in practice. Counselor helpful for producing defensible and usable research find-
researchers and counselors-as-practitioners routinely collect ings); (c) relies on qualitative and quantitative viewpoints,
and analyze qualitative and quantitative data as a necessary data collection, analysis, and inference techniques combined
part of their profession. Therefore, the purpose of this article according to the logic of mixed methods research to address
is threefold: (a) to demonstrate that regardless of philosophi- one’s research question(s); and (d) is cognizant, appreciative,
cal stance, collecting quantitative data via psychometrically and inclusive of local and broader sociopolitical realities,
sound quantitative instruments during the qualitative interview resources, and needs. (p. 129)
process enhances interpretations by helping researchers better
contextualize qualitative findings; (b) to explain the concept of If we take into account the integrative nature of counseling,
the mixed methods interview; and (c) to provide an example it is surprising that relatively few counseling researchers com-
demonstrating this strategy whereby a baseline was established bine or mix qualitative and quantitative data in their studies.
using a quantitative scale and normative data as a mixed re- Ray et al. (2011), who recently reviewed 4,457 articles from
search approach. 1998 to 2007 in 15 ACA division-affiliated journals, identi-
On the basis of definitions provided by 19 leading meth- fied only 171 mixed research articles, which represented only
odologists of mixed methods research, or more aptly named 3.84% of the total number of articles published in these jour-
as mixed research to denote the fact that more than methods nals. In fact, this finding is consistent with other researchers’
typically are mixed (e.g., philosophical assumptions and studies examining counseling journals that documented the
stances, research questions), Johnson, Onwuegbuzie, and lack of mixed research articles for either empirical research
Turner (2007) defined mixed research as articles or nonempirical research articles (e.g., theoretical/
conceptual articles; Hanson, Creswell, Plano Clark, Petska,
an intellectual and practical synthesis based on qualitative and & Creswell, 2005; Leech & Onwuegbuzie, 2011).
quantitative research; it is the third methodological or research Similar to other fields and disciplines, the low prevalence
paradigm (along with qualitative and quantitative research). rates of mixed research articles published in counseling
Rebecca K. Frels, Department of Counseling and Special Populations, Lamar University; Anthony J. Onwuegbuzie, Department
of Educational Leadership and Counseling, Sam Houston State University. The authors would like to acknowledge and thank John
Harris, Applied Research Consulting, and Michael Nakkula, University of Pennsylvania, for the use of the Match Characteristics
Questionnaire. The authors also thank Michael Karcher, University of Texas San Antonio, for his guidance in selecting the quantitative
measure and locating other resources on school-based mentoring. Correspondence concerning this article should be addressed
to Rebecca K. Frels, Department of Counseling and Special Populations, Lamar University, 223 Education Building, Beaumont, TX
77710 (e-mail: rebecca.frels@gmail.com).

© 2013 by the American Counseling Association. All rights reserved.

184 Journal of Counseling & Development  ■  April 2013  ■  Volume 91
Administering Quantitative Instruments With Qualitative Interviews

journals have occurred despite the exponential increase in a quantitative approach is used to address one part and a
the number of methodologically based mixed research ar- qualitative approach is used to address the other part); mixed
ticles that have been published in the literature (Ivankova & procedural/mixing research questions (i.e., narrow questions
Kawamura, 2010), including two handbooks (i.e., Tashakkori that direct the integration of the qualitative and quantitative
& Teddlie, 2003, 2010) and numerous books (i.e., Andrew strands of the study); combination research questions (i.e.,
& Halcomb, 2009; Bergman, 2008; Collins, Onwuegbuzie, at least one mixed research question combined with separate
& Jiao, 2010; Creswell & Plano Clark, 2010; Greene, 2007; quantitative and qualitative questions); independent research
Hess-Biber, 2010; Morse & Niehaus, 2009; Newman & Rid- questions (i.e., two or more research questions that are related,
enour, 2008; Onwuegbuzie, Jiao, & Bostick, 2004; Plowright, with each question not depending on the results of the other
2011; Teddlie & Tashakkori, 2009) on mixed research, as question[s]); dependent research questions (i.e., questions
well as guidelines written directly for counseling researchers that depend on the results stemming from addressing another
that were published in ACA’s flagship journal (i.e., Leech & question); predetermined research questions (i.e., questions
Onwuegbuzie, 2010). based on literature, practice, personal tendencies, and/or
These and other methodological works have demonstrated disciplinary considerations that are posed at the beginning
the utility of conducting mixed research. Moreover, although of the study); and emergent research questions (i.e., new or
quantitative research is particularly useful for “answering modified research questions that arise during the design, data
questions of who, where, how many, how much, and what is collection, data analysis, or interpretation phase).
the relationship between specific variables” (Adler, 1996, p. It is probable that many researchers might not conduct
5), it is not optimal for answering why and how questions. mixed research because of a lack of training or, as noted by
The converse is true for qualitative research. In contrast, Frels, Onwuegbuzie, Leech, and Collins (2012) and Onwueg-
mixed research can address both sets of questions within a buzie, Frels, Leech, and Collins (2011), a lack of pedagogical
single research study. Alternatively stated, mixed research has information in published form. Other reasons for the dearth of
been shown to be useful for addressing simultaneously both mixed research studies published in counseling journals might
quantitative-based questions that deal with prevalence rates be philosophical (i.e., researchers’ beliefs about the nature of
(i.e., descriptive research), relationships (i.e., correlational re- knowledge, objectivity–subjectivity dualism), axiological (i.e.,
search, causal-comparative, quasiexperimental research), and researchers’ beliefs about the role of values and ethics), or
cause-and-effect relationships (i.e., experimental research) ontological (i.e., researchers’ beliefs about the nature of reality).
and qualitative-based questions that lead to the examination In particular, at least some researchers mistakenly believe that
of local processes, experiences, and perceptions of individu- the philosophical assumptions and stances of their quantitative-
als such as counselees (e.g., biography, autobiography, life based research (e.g., postpositivism) or their qualitative-based
history, oral history, autoethnography, case study) and groups research (e.g., constructivism, critical theory) prevent them from
such as cultural groups (e.g., phenomenology, ethnography, mixing quantitative and qualitative approaches. However, as
grounded theory). demonstrated by Onwuegbuzie, Johnson, and Collins (2009),
Greene, Caracelli, and Graham (1989) identified five dif- the ontological, epistemological, and methodological assump-
ferent research purposes for mixing quantitative and qualita- tions and stances representing the major research paradigms
tive data: (a) triangulation (the intent is to seek convergence in do not prevent researchers from collecting and analyzing both
data); (b) complementarity (the intent is to measure overlap- quantitative and qualitative data—at least to some degree. Table
ping but different facets of a phenomenon); (c) development 1, which was created using Onwuegbuzie, Johnson, and Collins’s
(the intent is to help develop or inform the other method); (d) (2009) comparison of paradigms, displays major characteristics
initiation (the intent is to discover paradox and contradiction, associated with three qualitative-based paradigms (i.e., construc-
new perspectives of frameworks, and the recasting of ques- tivism, critical theory, and participatory), one quantitative-based
tions or results); and (e) expansion (the intent is to extend paradigm (i.e., postpositivism), and one mixed research-based
the breadth and range of inquiry). In addition, mixed research paradigm (i.e., pragmatism-of-the-middle) with respect to three
can be used to address a broader range of research questions axiomatic components (i.e., ontological, epistemological, and
than can monomethod studies (i.e., quantitative research alone methodological foundations). It can be seen from the table that
or qualitative research alone). For example, as identified by the philosophical assumptions and stances underlying postposi-
Plano Clark and Badice (2010), mixed research can be used tivism allow postpositivist researchers to utilize some qualitative
to address the following types of research questions: separate analysis techniques, especially those that yield frequency data
research questions (i.e., one or more quantitative research such as word count (i.e., counting particular times a word or
questions coupled with one or more qualitative research words are used) and classical content analysis (i.e., counting
questions); general overarching mixed research questions the codes).
(i.e., broad questions that are addressed using both quantita- At the other end of the research paradigmatic continuum,
tive and qualitative approaches); hybrid mixed issue research philosophical assumptions and stances associated with
questions (i.e., one question with two distinct parts such that qualitative inquiry—specifically constructivism—allow

Journal of Counseling & Development  ■  April 2013  ■  Volume 91 185

 Frels & Onwuegbuzie

Table 1
Contemporary Research Paradigms and Characteristics
Pragmatism-of-the-
Research Type Postpositivism Middle Critical Theory Constructivism Participatory

Ontology Social science inquiry Traditional dualisms Social, political, Multiple contradic- The mind and given
should be objective are rejected; high re- cultural, ethnic, tory, but equally world order are co-
gard for the influence racial, economic, valid accounts of the created through
of the inner world of and gender values same phenomenon subjective–objective
human experiences that evolve over time represent multiple reality
in action affect reality realities
Epistemology Researchers are Knowledge is based Transactional/ Co-created findings/ Experiential and
neutral, emotion- on the reality of the subjectivist; value- meaning; knowledge practical for practice;
ally detached, and world and constructed mediated findings is subjective and not co-created findings
should eliminate through experience; separable for the
biases; empirically justification comes knower
justify stated via warranted
hypotheses assertability
Methodology Generalizations are Thoughtful/ Use of a dialogue or Dialectical and impos- Political participation
time- and context- dialectical eclecti- dialectical approach sible to differentiate for collaborative
free; real causes cism and pluralism fully causes and action research;
of social scientific of methods and effects; uses induc- emphasizes the
outcomes can be perspectives; deter- tive reasoning; time- practical
determined reliably mines what works and context-free
and validly via and solves individual generalizations are
quantitative (and and social problems neither desirable nor
sometimes possible
qualitative) methods
Rhetoric Use of formal, Use of both imper- Use of critical Use of empathetic Use of language
impersonal, passive sonal passive voice/ discourse descriptions that are based on shared
voice and technical technical terminology informal, detailed, experiential context
terminology; focuses and empathetic, rich rich, and thick
on social laws and thick descrip-
tions
Qualitative analysis Some qualitative analy- All forms of qualitative All forms of qualitative All forms of qualitative All forms of qualitative
ses that generate analyses analyses analyses analyses
numbers as
part of the findings
(e.g., word count,
classical content
analysis)
Quantitative analysis All forms of descrip- All forms of descrip- Descriptive statis- Descriptive statistics; Descriptive statistics;
tive and inferential tive and inferential tics; most forms of some inferential inferential statistics
statistics for making statistics inferential statistics statistics for for both internal
external statistical for internal statistical internal statistical statistical
generalizationsa generalizationsb and generalization but generalizations and
external statistical not external statistical external statistical
generalizations generalization generalizations

Note. This table was created based on definitions found in The Sage Dictionary of Qualitative Inquiry (3rd ed.), by T. A. Schwandt,
2007, Thousand Oaks, CA: Sage; and information from “Toward a Philosophy of Mixed Data Analysis,” by A. J. Onwuegbuzie, R. B.
Johnson, and K. M. T. Collins, 2009, International Journal of Multiple Research Approaches, 3, pp. 122–123.
a
External statistical generalizations involve making generalizations, judgments, inferences, or predictions on data stemming from a
representative statistical (i.e., optimally random and large) sample of the population from which the sample was drawn (Onwuegbuzie,
Slate, Leech, & Collins, 2009). bInternal statistical generalizations involve making generalizations, judgments, inferences, or predictions
on data obtained from one or more representative or elite participants, such as key informants, subsample members, or politically
important cases, of the sample from which the participant(s) was selected (Onwuegbuzie, Slate, et al., 2009).

constructivist researchers (e.g., radical constructivists, cog- and stances underlying both critical theory and participatory
nitive constructivists, social constructivists/constructionists) paradigms allow critical theorist researchers and participatory
to use, at the very least, descriptive statistics (i.e., measures researchers, respectively, to use descriptive statistics and many
of central tendency [e.g., mean, median, mode, proportion]; forms of inferential statistics.
measures of variation/dispersion [e.g., variance, standard In the mixed research tradition, the philosophical assump-
deviation]; measures of position [e.g., percentile rank, z tions and stances that underlie mixed research represent the
score]; and measures of distributional shape [e.g., skewness, commensurability of paradigms. Johnson and Gray (2010)
kurtosis]). As also seen in Table 1, philosophical assumptions explained mixed methods thinking as considering how con-

186 Journal of Counseling & Development  ■  April 2013  ■  Volume 91

Administering Quantitative Instruments With Qualitative Interviews

flicting positions illuminate new learning. Onwuegbuzie, Therefore, quantitative researchers and qualitative re-
Johnson, and Collins (2009) described the coming together searchers can use mixed research techniques without contra-
of philosophical assumptions through 11 philosophical dicting their underlying research philosophical belief systems
stances. These stances fall within various points on a con- by conducting what they refer to as quantitative dominant
tinuum. As seen through the pragmatism-of-the-middle crossover mixed analysis and qualitative dominant crossover
paradigm in Table 1, the major philosophical assumptions mixed analysis, respectively.
and stances do not prevent researchers from using one or
more analysis types associated with one tradition (e.g., quan- Quantitative Dominant Crossover Mixed Analysis
titative analysis) to analyze data associated with a different According to Onwuegbuzie, Leech, and Collins (2011), who
tradition (e.g., qualitative data)—a concept Onwuegbuzie expanded the concept, quantitative dominant crossover mixed
and Combs (2010) called crossover mixed analyses. analysis is used when the researcher seeks to answer research
questions through a postpositivist (quantitative) stance while
Crossover Mixed Analyses also believing that qualitative data and analysis help address the
As a mixed research approach, crossover mixed analyses can research question(s) to a greater extent. This occurs at various
be used for the following: levels. At one end of the spectrum—the highest level—inte-
gration involves combining one or more sets of inferential
• Reduce: to condense the dimensionality of qualita- analyses with other types of qualitative analyses “for the
tive data/findings using quantitative analysis (e.g., purpose of integrated data reduction, integrated data display,
exploratory factor analysis of qualitative data) and/or data transformation, data correlation, data consolidation, data
quantitative data/findings using qualitative techniques comparison, data integration, warranted assertion analysis, and/
(e.g., thematic analysis of quantitative data; Onwueg- or data importation” (p. 377). At the lowest end of the spectrum,
buzie, 2003; Onwuegbuzie & Teddlie, 2003) quantitative dominant crossover mixed analysis is the combina-
• Display: to present visually qualitative and quantita- tion of one or more sets of inferential analyses with qualitative
tive results within the same display (Onwuegbuzie & analyses that generate some frequency data (e.g., word count)
Dickinson, 2008) because these data are closer to statistical data than are the data
• Transform: to convert quantitative data to be analyzed that would be generated by other qualitative analyses (e.g.,
qualitatively (i.e., qualitizing data) and/or qualitative data constant comparison analysis, discourse analysis). Regardless
into numerical codes that can be analyzed statistically of the level of integration, the quantitative strand would rep-
(i.e., quantitizing data; Tashakkori & Teddlie, 1998) resent the dominant strand, with the qualitative strand being
• Correlate: to associate qualitative data with quan- incorporated to address one or more of Greene et al.’s (1989)
titized data and/or quantitative data with qualitized five purposes for mixing (i.e., tri­angulation, complementarity,
data (Onwuegbuzie & Teddlie, 2003) development, initiation, and expansion), using one or more of
• Consolidate: to merge multiple data sets to create the nine crossover analysis types.
new or consolidated codes, variables, or data sets
(Onwuegbuzie & Teddlie, 2003) Qualitative Dominant Crossover Mixed Analysis
• Compare: to examine side-by-side qualitative and quan- According to Onwuegbuzie, Leech, and Collins (2011),
titative data/findings (Onwuegbuzie & Teddlie, 2003) qualitative dominant mixed analysis involves a philosophical
• Integrate: to incorporate qualitative and quantitative stance whereby the researcher assumes a (qualitative) con-
data/findings either into a coherent whole or into two structivist, critical theorist, or any stance that is associated
separate sets (i.e., qualitative and quantitative) of with the qualitative research paradigm and also believes that
coherent wholes (Onwuegbuzie & Teddlie, 2003) the addition of quantitative data and analysis would address
• Assert: to review all qualitative and quantitative data in more detail the research question(s). Building on this
to yield meta-inferences (Smith, 1997) idea, Ross and Onwuegbuzie (2011) categorized the array of
• Import data: to use follow-up findings from quali- established quantitative analysis techniques into the follow-
tative analysis to inform the quantitative analy- ing eight levels of complexity: Level 1, descriptive analyses
sis (e.g., qualitative contrasting case analysis, (e.g., measures of central tendency, dispersion, position);
qualitative residual analysis, qualitative follow- Level 2, univariate analyses (e.g., independent samples t test,
up interaction analysis, and qualitative internal dependent samples t test, one-way analysis of variance); Level
replication analysis) or follow-up f indings from 3, multivariate analyses (e.g., multiple analysis of variance,
quantitative analysis to inform the qualitative multiple analysis of covariance, discriminant analysis, canoni-
analysis (e.g., quantitative extreme case analysis, cal correlation analysis); Level 4, analyses of group mem-
quantitative negative case analysis; Onwuegbuzie bership (e.g., exploratory factor analysis, cluster analysis,
& Teddlie, 2003) correspondence analysis, multidimensional scaling); Level 5,

Journal of Counseling & Development  ■  April 2013  ■  Volume 91 187

 Frels & Onwuegbuzie

measurement techniques (e.g., confirmatory factor analysis, of counseling than for other fields. In fact, certain types of
item response theory); Level 6, analyses of time and/or space interviews, to a certain degree, can resemble the counseling
(e.g., autoregressive models, integrated models, moving av- interview process (e.g., Chenail, 1997; Ortiz, 2001). Thus, in
erage models, geocoding, geostatistics, cartography); Level many counseling specialties, including the field of marriage
7, multidirectional or multilevel analyses (e.g., structural and family therapy, interviews have been the most utilized
equation modeling, hierarchal linear modeling); and Level qualitative method (Gehart, Ratliff, & Lyle, 2001). As Chenail
8, multidirectional and multilevel analyses (e.g., multilevel (1997) declared, interviewing is a natural form of inquiry in
structural equation modeling, multilevel item response theory, the field of counseling because “it is so similar to the way in
multivariate hierarchical linear modeling). Thus, at the low- which counselors and therapists interact with their clients in
est level of integration, the qualitative dominant crossover therapy sessions” (Abstract).
mixed analysis would involve combining one or more sets We contend that the interview, as a natural mode of in-
of qualitative analyses with descriptive statistics (i.e., Level quiry, can be enhanced when researchers/interviewers collect
1 quantitative analysis). At a higher level of integration, the quantitative data alongside qualitative responses. Teddlie
qualitative dominant crossover mixed analysis would involve and Tashakkori (2009) referred to this strategy as within-
combining one or more sets of qualitative analyses with strategy mixed methods data collection. We call it a mixed
exploratory analysis techniques (i.e., Level 4 quantitative methods interview. Some examples of studies can be found
analysis), such as by subjecting the emergent themes to an in the literature wherein the researcher(s) developed and uti-
exploratory factor analysis (i.e., integrated data reduction; lized interview formats that contained both open-ended and
see Onwuegbuzie, 2003; Onwuegbuzie & Teddlie, 2003). closed-ended items (e.g., Brannen, 2005). However, our call
At the highest level of integration, the qualitative dominant is for an even more rigorous process of combining qualita-
crossover mixed analysis would involve combining one or tive open-ended interview questions with items from one or
more sets of qualitative analyses with inferential statistics more relevant (standardized) quantitative instruments (e.g.,
(i.e., Levels 2–3, 5–8). Whatever the level of integration, the Likert-format scales, rating scales) that possess adequate
qualitative strand would represent the dominant strand, with psychometric properties (i.e., adequate score reliability; ad-
the quantitative strand being used in an attempt to fulfill one equate score validity stemming from adequate content-related,
or more of Greene et al.’s (1989) five purposes for mixing. criterion-related, and construct-related validity), whenever
As Greene (2008) surmised, the combining of quantita- available, which allow the researcher(s) to contextualize
tive and qualitative analysis techniques “has not yet cohered further the qualitative interview responses.
into a widely accepted framework or set of ideas” (p. 14). Extracting standardized quantitative information—which
Furthermore, Bazeley (2010) concluded that “there are represents only Level 1 complexity on Ross and Onwueg-
surprisingly few published studies reporting results from buzie’s (2011) quantitative analysis continuum—alongside
projects which make more than very elementary use of the qualitative information from qualitative interviews enhances
capacity to integrate data and analyses using computers” (p. both representation and legitimation of the phenomenon of
434). Therefore, the concept of crossover analysis has great interest. Representation refers to the ability to extract an
potential for advancing the process of combining quantitative adequate amount of relevant information from each partici-
and qualitative data collection and data analysis techniques pant—optimally, under the conditions of saturation (Morse,
within the same framework. Indeed, Teddlie and Tashakkori 1995), particularly data saturation (i.e., when information
(2009) declared that this concept represents “one of the most emerges so repeatedly that the researcher can expect it and
fruitful areas for the further development of MM [mixed wherein the collection of more data appears to have no addi-
methods] analytical techniques” (p. 281). To this end, the tional interpretive worth; Sandelowski, 2008) and theoretical
purpose of the remainder of this article is to advance further saturation (i.e., when the researcher can assume that her or
one of the two components of crossover analyses, namely, his emergent theory is adequately developed to fit any future
the qualitative dominant crossover analysis. Specifically, we data collected; Sandelowski, 2008). Information gleaned from
illustrate how a qualitative dominant crossover analysis can the quantitative instrument(s) also lends clarity to the voice
enhance the quality of interpretation of interview data. of a participant. More specifically, in Greene et al.’s (1989)
typology, representation would be increased via enhanced
Qualitative Interviews complementarity, expansion, and development. Thus, by
incorporating additional sources of information, qualitative
Interviews represent one of the most common ways of col- researchers would obtain richer interpretations.
lecting data in qualitative research because they provide In contrast, legitimation refers to the validity of interpre-
opportunities for the researcher to collect rich and meaning- tations that stem from the interview data. Indeed, legitima-
making data (e.g., Roulston, 2010). Because of the therapeutic tion would be increased via the ability to compare and to
relationship and the role of a counselor, there is no doubt contrast the qualitative and quantitative data extracted from
that qualitative interviews are more relevant for the field the interview(s), again using Greene et al.’s (1989) triangu-

188 Journal of Counseling & Development  ■  April 2013  ■  Volume 91

Administering Quantitative Instruments With Qualitative Interviews

lation and initiation. And by increasing both representation pluralism, which refers to an epistemology that requires the
and legitimation by administering one or more standardized researcher to incorporate multiple epistemological perspec-
quantitative instruments, increased verstehen (i.e., under- tives. This philosophical stance lends itself to the use of a
standing) would ensue. What follows is a heuristic example crossover mixed analysis, and we combined epistemological
to illustrate, using a real study, the benefit of administering perspectives to include pragmatism-of-the-middle (Onwueg-
a standardized quantitative instrument as part of the qualita- buzie, Johnson, & Collins, 2009) and social constructionism
tive interview process. It is our belief that our exemplar also (Schwandt, 2000).
will serve as a model for understanding the mixed research
concepts previously discussed. Data Collection
To address the research questions, Frels (2010) conducted
Heuristic Example a multiple case study with 11 adult mentors (four men,
seven women), with ages ranging from 28 to 70 years and
The example we provide here was written by Frels (2010),
ethnicities of African American (n = 5), Hispanic (n = 2),
a professional school counselor (also the first author of
and White (n = 4). Each of these mentors was paired with
the present article), using a qualitative dominant crossover
a mentee such that the pairing involved one or each of the
mixed analysis within the context of a qualitative study
following two mentee–mentor pairings: same-gender versus
wherein interviews represented the main data collection
different-gender mentee–mentor pairings and same-ethnic
tool. The purpose of Frels’s (2010) study was to explore
versus different-ethnic mentee–mentor pairings.
selected mentors’ perceptions and experiences of the dyadic
Although many forms of data were collected with regard
mentoring relationship in school-based mentoring (SBM).
to all dyad interactions—including observations, descriptive
A second purpose was to build on the qualitative body of
case notes, reflexive data, and debriefing data—interviews
research (Spencer, 2004, 2007) for understanding roles,
represented the major data collection technique for explor-
purposes, approaches, and experiences of the relationship
ing the phenomenon of dyadic mentoring relationships. This
process with mentees (the dyadic relationship). The research
mode of inquiry resonated with the first author’s identity and
explored SBM as a type of helping relationship facilitated
relational approach in research as a professional school coun-
by a mentor, involving the untapped resources of the psy-
selor. Each mentor was interviewed separately on multiple
chotherapy literature and described by Spencer (2004),
occasions. Each interview, which lasted between 20 minutes
specifically, the dyadic relationship itself as the facilitator
and 60 minutes, was semistructured, with questions being
of change to affect both the mentor and the mentee. Frels’s
purposefully created to gain insight into the experience of the
research questions were as follows:
dyadic relationship. Examples of interview questions include
Research Question 1: What are the experiences and per- the following: What are your beliefs, thoughts, and opinions
ceptions of selected school-based mentors regarding about the purpose of mentoring? What words, phrases, or
roles, purposes, and approaches of mentoring within the images come to mind to describe the time you spend with
dyadic relationship with elementary school students? your mentee? When you feel challenged in your relationship,
Research Question 2: What are the differences and what are some thoughts or beliefs that help?
similarities in experiences and perceptions of selected In addition to the in-depth interviews, the 11 mentors
school-based mentors working with elementary school completed a standardized quantitative instrument, the 62-
students as a function of ethnicity of the mentor? item Match Characteristics Questionnaire (MCQ; Harris &
Nakkula, 2008), which measures the quality of matching
As noted by Plano Clark and Badice (2010), three ele- between mentors and mentees. The MCQ yielded good
ments are key in the focus of any study: the content area, psychometric properties, with the score reliability pertain-
the purpose, and the research questions. Even though ing to some of the subscales (e.g., Growth Focus, Support-
Frels’s (2010) research was a qualitative study, the research Seeking From the Mentee) ranging in the .90s. The MCQ
questions also might be considered as representing general subscale scores were used to contextualize the position of
overarching mixed research questions (i.e., broad questions each mentor relative to each other and also to obtain a richer
that are addressed using both quantitative and qualitative description of each of the 11 participants. The subscale
approaches) for driving the data collection methods. Con- scores informed both the ensuing cross-case analyses and
sequently, as explained by Plano Clark and Badice, research within-case analyses.
questions are inherently linked to environmental contexts
that include theories and beliefs. Therefore, we explored Data Analysis
belief systems and philosophies at the onset of the study Cross-case analysis. The following segments are excerpts
to recognize the lens from which data would be collected. from Frels’s (2010) report that provide examples of how the
As a result, the driving research paradigm was determined MCQ subscale scores were used to enhance the richness of
to be what Johnson (2011, 2012) labeled as dialectical interpretations stemming from the cross-case analysis:

Journal of Counseling & Development  ■  April 2013  ■  Volume 91 189

 Frels & Onwuegbuzie

To legitimate the metathemes and themes, scores from the Comparing the MCQ subscale scores with qualitative
MCQ (Harris & Nakkula, 2008) and selected subscales responses involved the following crossover analyses: inte-
were analyzed. Because norms are in the process of being grated data reduction, data transformation, data consolidation,
established, scores have been established as percentiles by the data comparison, data integration, and warranted assertion
authors of the MCQ (J. T. Harris, personal communication, analysis. Indeed, the selected mentors expressed satisfaction
June 2, 2010). On the whole, the selected mentors in my study (with the exception of the participant, Savannah, who scored
scored high on every subscale of the MCQ. (pp. 197–198) lower on the MCQ) in the interviews, and the use of the MCQ
  To explore relationship behaviors in each dyad and per- enriched this finding.
ceived program support (e.g., relating to Research Question Within-case analysis. The profiles of MCQ subscale scores
2), the following three subscales were examined for each played an important role in Frels’s (2010) decision to select
mentor: (a) program support subscale (i.e., the degree to Savannah for a follow-up, in-depth within-case analysis. The
which mentors feel that the program is providing effecting following excerpt from Frels’s (2010) report distinguishes a
training, supervision, and support); (b) support-seeking unique profile for Savannah:
broadscale (i.e., the degree to which mentors feel that their
mentees seek their support in relation to personal issues and Savannah’s profile on the MCQ indicated an equal to or higher
academics); and (c) mentor satisfaction subscale (i.e., the score than the 75th percentile on three (comfort, fun focus, future
degree to which mentors feel that their match is growing outlook) of the 10 subscales utilized to measure relationship
stronger and producing good results for the mentee). characteristics. On three subscales (closeness, character develop-
  Interestingly, Savannah and Chad [pseudonyms of two ment, relating focus), Savannah scored above the average range
participants of the study] scored lowest in the area of support- but below the 75th percentile. Additionally, Savannah scored
seeking as they did in the area of sharing. In addition, Savannah particularly low in the areas of sharing (lower than average) and
scored below the 50th percentile in all three categories: program satisfaction (at the 25th percentile). Figure 32 [Figure 1 in the
support, support-seeking behavior, and mentor satisfaction. Of present article] depicts the scores of Savannah as they relate to
the 11 mentors, seven scored above the 50th percentile for all the MCQ averages, the 75th percentile, and the 25th percentile.
three categories. Two of 11 mentors scored higher than the 75th Interestingly, Savannah scored highest in the area of focusing
percentile for two of the three categories; and only one mentor on the future (Sf = 96 < 75th percentile = 80). As seen in [Figure
(Savannah) scored low for all three categories. (pp. 200–201) 1 in the present article], Savannah’s profile on the MCQ was
100 —

90 —

80 —

70 —

60 — Savannah

Average
50 —
75th percentile

40 — 25th percentile
s

s
for t

s
s

n
s
t
ess

men

u
u

u
ocu

ocu

ctio
Foc

Foc
Foc

Foc
com
sen

ic F

gF

isfa
p
velo

Fun

wth
r ing

look
Clo

Dis

atin

Sat
dem

Gro
r De

Sha

Out
Rel
Aca
e

ure
ract

Fut
Cha

Figure 1
Averages of Selected Subscales From the Match Characteristics Questionnaire
and the Profile of Savannah
Note. Adapted from “The Experiences and Perceptions of Selected Mentors: An Exploratory Study of the Dyadic Relationship in
School-Based Mentoring,” by R. K. Frels, 2010, unpublished doctoral dissertation, pp. 212–213. Copyright 2010 by R. K. Frels.

190 Journal of Counseling & Development  ■  April 2013  ■  Volume 91

Administering Quantitative Instruments With Qualitative Interviews

very high or very low on various subscales. Hence, her profile mentarity—to measure overlapping but different facets of a
is deemed: highly fluctuating. Table 27 [Table 2 in the present phenomenon (Greene et al., 1989)—with the case of Savan-
article] provides statements (i.e., qualitative data that support the nah. The constant comparison analysis of qualitative data
MCQ responses). (p. 211) yielded the theme of Too Many Questions, which seemed to
hinder the dyadic relationship. Frels presented the subscale
As a result of the analysis, the table described in the excerpt Satisfaction in relationship to the theme Too Many Questions
(see Table 2) was a reference point that aligned the quantitative to contextualize how the use of questions was inherent in
instrument with some of the qualitative findings. For example, Savannah’s relating style. For example, Savannah described
as seen in Table 2, Savannah scored in the high fourth quartile her own disappointment of the mentoring experience through
for the MCQ subscale Future Outlook Focus. This technique questions during an interview: “Why are you [myself] here?
of correlating scores with qualitative information involved . . .you know . . . why did they [the mentees] want you to
the following crossover analyses: integrated data reduction, come? And was it their idea? Was it their parents’ idea? Was
integrated data display, data transformation, data correlation, it their teacher’s idea?” Thus, with the use of Greene et al.’s
data consolidation, data comparison, data integration, and (1989) purposes for mixing as a frame for a visual display,
warranted assertion analysis. During the interview, Savannah data integration in the qualitative dominant crossover mixed
disclosed that she was unhappy with the progress that she and analysis was evident.
her mentee were making. Subsequently, Savannah scored in
the lower third quartile of the MCQ subscale Satisfaction. As
seen in Table 2, the data were integrated by displaying this
Conclusion
MCQ quantitative score aligned with an example quotation As we have shown in this article, supplementing open-ended
from the interview. Savannah’s high expectations and unre- interview responses with quantitative data from one or more
alistic goals resulted in her frustration and her decision to psychometrically sound (standardized) quantitative instru-
discontinue mentoring (Frels & Onwuegbuzie, 2012a, 2012b). ments can increase the rigor of qualitative studies; this practice
This concept resonates with other SBM literature (Karcher, is consistent with many philosophical paradigms. In addi-
Herrera, & Hansen, 2010), whereby goal-oriented interactions tion, by recognizing the value of crossover mixed analyses,
(e.g., focusing on the future) often are not sufficient indicators researchers might view philosophical integration much like
of relationship closeness in SBM. Furthermore, revealing the how they would view the concept of theoretical integration
relationship between Savannah’s intent to leave mentoring in counseling. Oftentimes, counselors adhere to one guiding
and the MCQ Satisfaction subscale provided evidence of theory, which can be integrated with points in common with
triangulation or convergence in data. other theoretical concepts, including underlying philosophy,
Finally, with reference to Greene et al.’s (1989) five differ- values, and data collection (Kottler & Montgomery, 2011).
ent research purposes for mixing quantitative and qualitative Because incorporating information from standardized
data, themes from the constant comparison analysis of inter- quantitative instruments into the analysis of qualitative
view data and the MCQ scores were mapped and identified interview data represents the use of quantitative analysis
with one or more purposes. This type of data correlation techniques that are classified only as Level 1 complexity
map can provide further evidence of how a researcher can (Ross & Onwuegbuzie, 2011), this strategy should not con-
integrate qualitative findings with a quantitative instrument. tradict the philosophical assumptions and stances of any of
For example, Frels (2010) presented the purpose of comple- the major qualitative-based research paradigms (e.g., con-

Table 2
Correlating Match Characteristics Questionnaire Subscale Scores
With Qualitative Statements
Dimension/Subscale Quartile Examples From Savannah’s Interview
Academic Focus Low fourth “I mean you have to have something—like if her teacher wanted her to work on fractions—she
could have sent maybe something for her to work on with us, to work on together.”
Relating Focus Low fourth “Well, we played a couple of games but I think talking more connecting . . . because when we
played the game—we didn’t talk so much.”
Future Outlook Focus High fourth “And um, I [mentor] . . . really studied. I worked hard but I did well. And I thought ‘wow—wow’
and I just want to encourage her ’cause things didn’t come easy to me; they don’t come easy
to her. And I wanted to give her a head start. Don’t wait till you’re in college or until after your
kids are born to learn how to study.”
Satisfaction Low third “The fact that she [mentee] was shy and the fact that she still struggles in school, I wanted to
help but those two things really kind of made me decide, you know, on what I did [to quit].
Some days, it was hard.”
Note. Adapted from “The Experiences and Perceptions of Selected Mentors: An Exploratory Study of the Dyadic Relationship in
School-Based Mentoring,” by R. K. Frels, 2010, unpublished doctoral dissertation, pp. 212–213. Copyright 2010 by R. K. Frels.

Journal of Counseling & Development  ■  April 2013  ■  Volume 91 191

 Frels & Onwuegbuzie

structivism, critical theory, participatory). Thus, even though Collins, K. M. T., Onwuegbuzie, A. J., & Jiao, Q. G. (Vol. Eds.).
supplementing qualitative interview data with quantitative (2010). The research on stress and coping in education series:
data—what we call a mixed methods interview—leads to Vol. 5. Toward a broader understanding of stress and coping:
a mixed analysis, the resultant mixed analysis would be Mixed methods approaches. Greenway, CT: Information Age
qualitatively dominant. Furthermore, as Guba and Lincoln Publishing.
(2011) wrote, Creswell, J. W., & Plano Clark, V. L. (2010). Designing and conduct-
ing mixed methods research (2nd ed.). Thousand Oaks, CA: Sage.
Are paradigms commensurable? Is it possible to blend ele- Frels, R. K. (2010). The experiences and perceptions of selected
ments of one paradigm into another, so that one is engaging mentors: An exploratory study of the dyadic relationship in
in research that represents the best of both worldviews? The school-based mentoring (Unpublished doctoral dissertation).
answer, from our perspective, has to be a cautious yes. This is Sam Houston State University, Huntsville, TX.
so if the models (paradigms, integrated philosophical systems) Frels, R. K., & Onwuegbuzie, A. J. (2012a). The experiences of
share axiomatic elements that are similar, or that resonate selected mentors: A cross-cultural examination of the dyadic
strongly between them. (p. 117) relationship in school-based mentoring. Mentoring & Tutoring:
Partnership in Learning, 20, 1–26. doi:10.1080/13611267.201
Therefore, our call for an even more rigorous process of 2.679122
combining qualitative open-ended interview questions with Frels, R. K., & Onwuegbuzie, A. J. (2012b). Principles of play:
items from standardized quantitative instruments, via a mixed A dialogical comparison of two case studies in school-based
methods interview, represents the blending of elements of mentoring. International Journal of Play Therapy, 21, 131–148.
one paradigm into another that provides qualitative research- doi:10.1037/a0028536
ers from the field of counseling the best of both worldviews. Frels, R. K., Onwuegbuzie, A. J., Leech, N. L., & Collins, K. M..
Most important, the collection of quantitative data during the (2012). Challenges to teaching mixed research courses. Journal
qualitative interview process allows researchers to compare of Effective Teaching, 12, 23–44.
each interviewee with extant normative data, including inter- Gehart, D. R., Ratliff, D. A., & Lyle, R. R. (2001). Qualitative
national norms, national norms, regional norms, local norms, research in family therapy: A substantive and methodological
and relevant cultural norms. Thus, we encourage qualitative review. Journal of Marital and Family Therapy, 27, 261–270.
researchers, whenever appropriate, to administer one or more doi:10.1111/j.1752-0606.2001.tb01162.x
quantitative instruments that tap the construct of interest to Greene, J. C. (2007). Mixed methods in social inquiry. San Francisco,
increase verstehen. CA: Jossey-Bass.
Greene, J. C. (2008). Is mixed methods social inquiry a distinctive
References methodology? Journal of Mixed Methods Research, 2, 7–22.
doi:10.1177/1558689807309969
Adler, L. (1996). Qualitative research of legal issues. In D. Schimmel Greene, J. C., Caracelli, V. J., & Graham, W. F. (1989). Toward a
(Ed.), Research that makes a difference: Complementary methods conceptual framework for mixed-method evaluation designs.
for examining legal issues in education (NOLPE Monograph Educational Evaluation and Policy Analysis, 11, 255–274.
Series No. 56, pp. 3–31). Topeka, KS: National Organization doi:10.3102/01623737011003255
on Legal Problems of Education. Guba, E. G., & Lincoln, Y. S. (2011). Paradigmatic controversies,
American Counseling Association. (2005). ACA code of ethics. contradictions, and emerging confluences, revisited. In N. K.
Alexandria, VA: Author. Denzin & Y. S. Lincoln (Eds.), The Sage handbook of qualita-
Andrew, S., & Halcomb, E. J. (Eds.). (2009). Mixed methods re- tive research (4th ed., pp. 97–128). Thousand Oaks, CA: Sage.
search for nursing and the health sciences. Chichester, England: Hanson, W. E., Creswell, J. W., Plano Clark, V. L., Petska, K. S.,
Wiley-Blackwell. & Creswell, J. D. (2005). Mixed methods research designs in
Bazeley, P. (2010). Computer-assisted integration of mixed methods counseling psychology. Journal of Counseling Psychology, 52,
data sources and analysis. In A. Tashakkori & C. Teddlie (Eds.), 224–235. doi:10.1037/0022-0167.52.2.224
Sage handbook of mixed methods in social and behavioral Harris, J. T., & Nakkula, M. J. (2008). Match Characteristic Ques-
research (2nd ed., pp. 431–467). Thousand Oaks, CA: Sage. tionnaire (MCQ). Unpublished measure, Harvard Graduate
Bergman, M. (Ed.). (2008). Advances in mixed methods research: School of Education.
Theories and applications. Thousand Oaks, CA: Sage. Hess-Biber, S. N. (2010). Mixed methods research: Merging theory
Brannen, J. (2005). Mixing methods: The entry of qualitative and with practice. New York, NY: Guilford Press.
quantitative approaches into the research process. International Ivankova, N. V., & Kawamura, Y. (2010). Emerging trends in the
Journal of Social Research Methodology, 8, 173–184. utilization of integrated designs in the social, behavioral, and
Chenail, R. J. (1997). Interviewing exercises: Lessons from family health sciences. In A. Tashakkori & C. Teddlie (Eds.), Sage
therapy. The Qualitative Report, 3. Retrieved from http://www. handbook of mixed methods in social and behavioral research
nova.edu/ssss/QR/QR3-2/chenail.html (2nd ed., pp. 581–611). Thousand Oaks, CA: Sage.

192 Journal of Counseling & Development  ■  April 2013  ■  Volume 91

Administering Quantitative Instruments With Qualitative Interviews

Johnson, R. B. (2011). Dialectical pluralism: A metaparadigm to help Onwuegbuzie, A. J., Jiao, Q. G., & Bostick, S. L. (2004). Library
us hear and ‘‘combine” our valued differences. In S. J. Hesse-Biber anxiety: Theory, research, and applications (Research Methods
(Chair), Addressing the credibility of evidence in mixed methods in Library and Information Studies, No. 1). Lanham, MD:
research: Questions, issues and research strategies. Symposium Scarecrow Press.
conducted at the meeting of Seventh International Congress of Onwuegbuzie, A. J., Johnson, R. B., & Collins, K. M. T. (2009). A
Qualitative Inquiry, University of Illinois at Urbana-Champaign. call for mixed analysis: A philosophical framework for combining
Johnson, R. B. (2012). Dialectical pluralism and mixed re- qualitative and quantitative. International Journal of Multiple
search. American Behavioral Scientist, 56, 751–754. Research Methods, 3, 114–139.
doi:10.1177/0002764212442494 Onwuegbuzie, A. J., Leech, N. L., & Collins, K. M. T. (2011). Toward
Johnson, R. B., & Gray, R. (2010). A history of philosophical and a new era for conducting mixed analyses: The role of quantitative
theoretical issues for mixed methods research. In A. Tashakkori dominant and qualitative dominant crossover mixed analyses.
& C. Teddlie (Eds.), Sage handbook of mixed methods in social In M. Williams & W. P. Vogt (Eds.), The Sage handbook of in-
and behavioral research (2nd ed., pp. 581–611). Thousand novation in social research methods (pp. 353–384). Thousand
Oaks, CA: Sage. Oaks, CA: Sage.
Johnson, R. B., Onwuegbuzie, A. J., & Turner, L. A. (2007). Toward a Onwuegbuzie, A. J., Slate, J. R., Leech, N. L., & Collins, K. M. T.
definition of mixed methods research. Journal of Mixed Methods (2009). Mixed data analysis: Advanced integration techniques.
Research, 1, 112–133. doi:10.1177/1558689806298224 International Journal of Multiple Research Approaches, 3, 13–33.
Karcher, M. J., Herrera, C., & Hansen, K. (2010). “I dunno, what do Onwuegbuzie, A. J., & Teddlie, C. (2003). A framework for analyzing
you wanna do?”: Testing a framework to guide mentor training data in mixed methods research. In A. Tashakkori & C. Teddlie
and activity selection. New Directions for Youth Development, (Eds.), Handbook of mixed methods in social and behavioral
126, 51–69. doi:10.1002/yd.349 research (pp. 351–383). Thousand Oaks, CA: Sage.
Kottler, J. A., & Montgomery, M. J. (2011). Theories of counseling Ortiz, S. M. (2001). How interviewing became therapy for
and therapy: An experiential approach. Thousand Oaks, CA: Sage. wives of professional athletes: Learning from a seren-
Leech, N. L., & Onwuegbuzie, A. J. (2010). Guidelines for con- dipitous experience. Qualitative Inquiry, 7, 192–220.
ducting and reporting mixed research in the field of counseling doi:10.1177/107780040100700204
and beyond. Journal of Counseling & Development, 88, 61–69. Plano Clark, V. L., & Badice, M. (2010). Research questions in mixed
doi:10.1002/j.1556-6678.2010.tb00151.x methods research. In A. Tashakkori & C. Teddlie (Eds.), Sage
Leech, N. L., & Onwuegbuzie, A. J. (2011). Mixed research handbook of mixed methods in social and behavioral research
in counseling: Trends in the literature. Measurement and (2nd ed., pp. 275–304). Thousand Oaks, CA: Sage.
Evaluation in Counseling Development, 44, 169–180. Plowright, D. (2011). Using mixed methods: Frameworks for an
doi:10.1177/0748175611409848 integrated methodology. Thousand Oaks, CA: Sage.
Morse, J. M. (1995). The significance of saturation. Qualitative Powell, H., Mihalas, S., Onwuegbuzie, A. J., Suldo, S., & Daley, C. E.
Health Research, 5, 147–149. doi:10.1177/104973239500500201 (2008). Mixed methods research in school psychology: A mixed
Morse, J. M., & Niehaus, L. (2009). Mixed method design: Principles methods investigation of trends in the literature. Psychology in
and procedures. Walnut Creek, CA: Left Coast Press. the Schools, 45, 291–309. doi:10.1002/pits.20296
Newman, I., & Ridenour, C. R. (2008). Mixed methods research. Ray, D. C., Hull, D. M., Thacker, A. J., Pace, L. S., Swan, K. L.,
Chicago, IL: Southern Illinois University Press. Carlson, S. E., & Sullivan, J. M. (2011). Research in counseling:
Onwuegbuzie, A. J. (2003). Effect sizes in qualitative research: A 10-year review to inform practice. Journal of Counseling & De-
A prolegomenon. Quality & Quantity: International Journal velopment, 89, 349–359. doi:10.1002/j.1556-6678.2011.tb00099.x
of Methodology, 37, 393–409. doi:10.1023/A:1027379223537 Ross, A., & Onwuegbuzie, A. J. (2011, February). Complexity of
Onwuegbuzie, A. J., & Combs, J. P. (2010). Emergent data analy- quantitative analyses used in mixed research articles from the field
sis techniques in mixed methods research: A synthesis. In A. of mathematics education. Paper presented at the annual meeting
Tashakkori & C. Teddlie (Eds.), Handbook of mixed methods in of the Eastern Educational Research Association, Sarasota, FL.
social and behavioral research (2nd ed., pp. 397–430). Thousand Roulston, K. (2010). Considering quality in qualita-
Oaks, CA: Sage. tive interviewing. Qualitative Research, 10, 199–228.
Onwuegbuzie, A. J., & Dickinson, W. B. (2008). Mixed methods doi:10.1177/1468794109356739
analysis and information visualization: Graphical display for Sandelowski, M. (2008). Theoretical saturation. In L. M. Given
effective communication of research results. Qualitative Report, (Ed.), The Sage encyclopedia of qualitative methods (Vol. 1, pp.
13, 204–225. Retrieved from http://www.nova.edu/ssss/QR/ 875–876). Thousand Oaks, CA: Sage.
QR13-2/onwuegbuzie.pdf Schwandt, T. A. (2000). Three epistemological stances for qualitative
Onwuegbuzie, A. J., Frels, R. K., Leech, N. L., & Collins, K. M. inquiry: Interpretivism, hermeneutics, and social constructio-
T. (2011). A mixed research study of mixed research courses: nism. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of
Experiences and perceptions of instructors. International Journal qualitative research (2nd ed., pp. 189-215). Thousand Oaks,
of Multiple Research Approaches, 5, 169–202. CA: Sage.

Journal of Counseling & Development  ■  April 2013  ■  Volume 91 193

 Frels & Onwuegbuzie

Schwandt, T. A. (2007). The Sage dictionary of qualitative inquiry Tashakkori, A., & Teddlie, C. (1998). Applied Social Research Meth-
(3rd ed.). Thousand Oaks, CA: Sage ods Series: Vol. 46. Mixed methodology: Combining qualitative
Smith, M. L. (1997). Mixing and matching: Methods and models. In and quantitative approaches. Thousand Oaks, CA: Sage.
J. C. Greene & V. J. Caracelli (Eds.), Advances in mixed-method Tashakkori, A., & Teddlie, C. (Eds.). (2003). Handbook of mixed
evaluation: The challenges and benefits of integrating diverse methods in social and behavioral research. Thousand Oaks,
paradigms (New Directions for Evaluation No. 74, pp. 73–85). CA: Sage.
San Francisco, CA: Jossey-Bass. Tashakkori, A., & Teddlie, C. (Eds.). (2010). Sage handbook of mixed
Spencer, R. (2004). Studying relationships in psychotherapy: An methods in social and behavioral research (2nd ed.). Thousand
untapped resource for youth mentoring. New Directions for Youth Oaks, CA: Sage.
Development, 103, 31–42. doi:10.1002/yd.89 Teddlie, C., & Tashakkori, A. (2009). Foundations of mixed
Spencer, R. (2007). “It’s not what I expected”: A qualitative study methods research: Integrating quantitative and qualitative
of youth mentoirng relationship failures. Journal of Adolescent techniques in the social and behavioral sciences. Thousand
Research, 22, 331–354. doi:10.1177/0743558407301915 Oaks, CA: Sage.

194 Journal of Counseling & Development  ■  April 2013  ■  Volume 91