A call for more mixed methods in sport management research

  • Published on
    28-Oct-2016

  • View
    216

  • Download
    0

Embed Size (px)

Transcript

<ul><li><p>A call for more mixed methods in sport management research</p><p>Andy Rudd a,*, R. Burke Johnson b,1</p><p>a Florida State University, Department of Sport Management, Recreation Management, &amp; Physical Education, Tallahassee, FL 32306-4280, United StatesbUniversity of South Alabama, College of Education, BSET, 3700 UCOM, Mobile, AL 36688, United States</p><p>Many social and behavioral science researchers have promoted the use of mixed methods to more effectively answerresearch questions (Brewer &amp; Hunter, 1989; Johnson &amp; Onwuegbuzie, 2004; Tashakkori &amp; Teddlie, 1998, 2003). Such anapproach has generally been dened as the combining of at least one quantitative method and one qualitative method (e.g.,Hanson, Creswell, Plano Clark, Petska, &amp; Creswell, 2005; Jick, 1979; Maxwell &amp; Loomis, 2003). Combining quantitative andqualitative data in a single study can be benecial in a variety of ways. For example, the researcher can triangulate whichinvolves combining quantitative and qualitative methods to produce a set of data that has complementary strengths andnonoverlapping weaknesses (Brewer &amp; Hunter, 1989; Johnson &amp; Onwuegbuzie, 2004; Johnson &amp; Turner, 2003; Tashakkori &amp;Teddlie, 1998).</p><p>This concept of combining approaches for complementary strengths and non overlappingweaknesses has been called thefundamental principle of mixed research (Johnson &amp; Turner, 2003). The idea is to strategically select a mixture ofquantitative and qualitative approaches thatwill effectively cover the objective or set of objectives of a research study and todo it in way that eliminates overall study design weaknesses. According to Onwuegbuzie and Johnson (2006) thisfundamental principle can be adapted to multiple purposes including initiation (discovering contradictions), expansion(attaining a deeper and broader understanding), complementarity (examining overlapping parts of a phenomenon), anddevelopment (using results from one method to inform the use of a second method) (see also, Greene, Caracelli, &amp; Graham,1989).</p><p>The potential benets of using mixed methods approaches has stimulated its adoption in a variety of elds includingsociology, nursing, psychology, management, health sciences, evaluation, and education (Tashakorri and Teddlie, 2003). Yet,despite the popularity and strong advocacy for combining quantitative and qualitative approaches, few mixed methodsstudies can be found amid extant sport management research. As evidence, empirical articles were reviewed by the rst</p><p>Sport Management Review 13 (2010) 1424</p><p>A R T I C L E I N F O</p><p>Article history:</p><p>Received 18 September 2008</p><p>Received in revised form 1 April 2009</p><p>Accepted 18 June 2009</p><p>Keywords:</p><p>Mixed methods</p><p>Cause and effect</p><p>Causal explanation</p><p>Causal description</p><p>A B S T R A C T</p><p>Despite the popularity and strong advocacy for combining quantitative and qualitative</p><p>methods, fewmixedmethods approaches are found in the sportmanagement research. As a</p><p>result, this article examines the frequency with which mixed methods research has been</p><p>used in recent sportmanagement research, anddemonstratesways inwhichmixedmethods</p><p>can help improve the validity of research ndings in sport management related topics.</p><p>Because research in sportmanagement often is concernedwith causal questions, this article</p><p>provides mixed methods designs for improving causal inference. Examples are provided</p><p>from three areas of sport management research, including marketing, organizational</p><p>behavior, andnance. The designs that are provided are based on themixedmethods design</p><p>dimensions of time order and priority of quantitative and qualitative data.</p><p> 2009 Sport Management Association of Australia and New Zealand. Published byElsevier Ltd. All rights reserved.</p><p>* Corresponding author. Tel.: +1 850 645 6883.</p><p>E-mail addresses: rudd@coe.fsu.edu (A. Rudd), bjohnson@usouthal.edu (R.B. Johnson).1 Tel.: +1 251 380 2861.</p><p>Contents lists available at ScienceDirect</p><p>Sport Management Review</p><p>journal homepage: www.e lsev ier .com/ locate /smr</p><p>1441-3523/$ see front matter 2009 Sport Management Association of Australia and New Zealand. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.smr.2009.06.004</p></li><li><p>author for the presence of mixed methods approaches in three major sport management journals: Journal of SportManagement (JSM), International Journal of Sport Management (IJSM), and Sport Management Review (SMR) during theperiod of 20002007.2 The analysis identied the following number of mixedmethods articles: JSM (10), IJSM (15), and SMR(2) (Rudd, 2007). Comparatively, Quarterman et al. (2006) reviewed 299 articles (conceptual and empirical) published in theJournal of Sport Management from 1987 to 2004. From their analysis, only 6 of the 299 articles (2%) employed a mixedmethods approach, 27 out 299 (9%) were qualitative, 165 of the 299 (55%) were quantitative, and 101 of the 299 (34%) wereconceptual. As well, Barber, Parkhouse, and Tedrick (2001) assayed 42 empirical articles published in the Journal of SportManagement during the period of 19911995 and found the majority of articles to be survey research (73%). Other types ofmethods employed included content analysis (19%) and mixed methods (9%). Overall, these analyses suggest that when itcomes to empirical studies, the majority in sport management employ quantitative monomethods (a single method).</p><p>Additionally, although there aremany different types ofmixedmethods designs (Creswell &amp; Plano Clark, 2007; Johnson &amp;Onwuegbuzie, 2004; Teddlie &amp; Tashakkori, 2006), many of the mixed methods articles uncovered in our analysis involvedlimited or weak use of mixed methods. According to Johnson and Christensen (2008), research methodology can be viewedon a continuum ranging from monomethods (far left of the continuum) to fully mixed studies (far right of the continuum).The majority of the mixed methods studies observed in sport management journals would arguably fall to the left on theresearch continuum, representing relativelyweak use of combining quantitative and qualitativemethods. For example, 10 ofthe 15 mixed methods articles found in IJSM simply involved a questionnaire with open and closed questions. Otherexamples of weak to moderate mixed methods included eight articles involving content analysis (within JSM and IJSM) andve articles across JSM, IJSM, and SMR that used mixed methods for instrument development (i.e., using qualitative datafrom interviews or focus groups to develop items for a questionnaire).</p><p>In sum, the results from our analysis as well others (Barber et al., 2001; Quatermen et al., 1999) suggest that few employmixed methods approaches in sport management and when they do, they are not strongly mixed designs (i.e., a substantialamount of both quantitative and qualitative strategies and data in single studies). The purpose of this article, then, is todemonstrate additional, more fully mixed methods approaches that can be useful to the sport management researcher. Inparticular, given that much research in sport management is concerned with causal questions, our article focuses onapplyingmixedmethods to studies dealingwith causation. Becausewe are focusing on the question of causation, the type ofmixed methods research we are discussing in this article generally involved complementing quantitative research withqualitative approaches. The reason is that quantitative epistemologies are explicitly interested in identifying nomological(causal) relationships in their literatures.</p><p>For example, in the area of sportmarketing, researchers are often interested in the factors that affect purchasing behaviorof sport fans (e.g., Kwon &amp; Armstrong, 2002; Kwon, Trail, &amp; James, 2007; Trail, Fink, &amp; Anderson, 2003). Or, concerningorganizational behavior, studies have examined variables that inuence organizational commitment or turnover intention(e.g., Cunningham &amp; Sagas, 2006; Kent &amp; Chelladurai, 2001; Whisenant, 2005). As well, in the domain of nance andeconomics, researchers commonly ask causal questions related to economic impact. For example, a professional sport teamseffect on public goods (Johnson, Mondello, &amp;Whitehead, 2007) or a star players effect on fan attendance (DeSchriver, 2007).To probe cause and effect relationships many of these studies employ correlational or causal modeling techniques ratherthan experimental or quasi-experimental designs. This is presumably because it is difcult to control or manipulatenaturally occurring causal variables such as team identity, the location of a professional sport team, or the procurement ofstar players.3</p><p>Regardless of the method one uses to ascertain cause and effect, an important consideration is the distinction betweencausal description and causal explanation. The former simply allows one to suggest that there is a causal relationship orcovariance between the independent and dependent variable. The latter involves an explanation of the causal mechanismsresponsible for a particular causal relationship as well as the conditions under which the relationship holds (Cook, 2002;Shadish, Cook, &amp; Campbell, 2002) (see a more extensive explanation later in this article). Obtaining this explanatoryknowledge allows researchers to make more accurate conclusions (increases internal and external validity) and providesclearer applicability and utility to practitioners.</p><p>Although causal or structural equation models (SEM) take into consideration mediating and moderating variables thatmay help explain a causal relationship, these models may fail to capture additional important mediating or moderatingfactors. This problem is known as specication error which refers to misidentifying or omitting important variables that are</p><p>2 It is acknowledged that mixed method approaches have been employed by researchers since at least the latter half of the 20th century (Brewer &amp;</p><p>Hunter, 1989; Greene et al., 1989). Therefore, it is possible that the journals we reviewed carry articles employing mixed-method approaches prior to the</p><p>21st century. However, we felt a review of recent empirical articles over the last 7 yearswould provide a reasonable estimation of how oftenmixedmethods</p><p>are used in sport management research.3 Some of the studies identied employed correlational or statistical modeling techniques and use language such as relationship rather than effect.</p><p>We argue thatmany of these studies are essentially interested in a cause and effect relationship rather than a correlation. For example, Kent and Chelladurai</p><p>(2001) studied the correlation between transformational leadership and organizational commitment. A personal communicationwith Kent revealed that</p><p>he and his colleague were interested in a cause and effect relationship but refrained from such language given the lack of an experimental design and the</p><p>ability to control or manipulate the independent variables (A. Kent, personal communication, November 16, 2007). We surmise that many other studies</p><p>may operate under a similar rationale, and in fact, some research methods textbooks sometimes encourage writers to do so. We recommend that authors</p><p>use language appropriate to the purpose of the study while simultaneously admitting and discussing the weaknesses of the research design.</p><p>A. Rudd, R.B. Johnson / Sport Management Review 13 (2010) 1424 15</p></li><li><p>involved in a causal relationship (Shadish et al., 2002). Specication error also occurs when the researcher incorrectlyidenties the direction of a causal relationship or mistakes correlation for causation (Shadish et al.).</p><p>As an example of specication error, Kwon et al. (2007) examined the mediating effect of perceived value in therelationship between team identication and intent to purchase collegiate team-licensed apparel. In their discussion theyacknowledged that they were surprised not to nd a direct signicant relationship between team identication andpurchase intention. They surmised that the failed relationship may have been related to including the price of the t-shirtwhereas past studies had not. The researchers theorized that perhaps price sensitivity had a moderating effect on purchaseintention. Kwon et al.s study is a prime example of how the addition of qualitative data may strengthen causal explanation.Specically, qualitative methods including focus groups, one-on-one interviews, and eld observations could have beenused to expand their understanding of various unanticipated moderating or mediating factors. This article will thereforeprovide specic strategies for adding qualitative data to quantitative data to increase causal explanation in cause and effectrelated studies. However, it is important to note that we do not intend to suggest that the addition of qualitative data canovercome all of the weaknesses involved in making causal inferences with correlational techniques like SEM. Rather, thepoint is to show how the addition of qualitative data can expand and improve ones understanding of causal relationships.Additionally, we advocate the use of mixed method approaches for not only correlational techniques but also experimentaldesigns (see Cook, 2002; Shadish et al., 2002).</p><p>Before describing the mixed methods approaches, it is important to familiarize the reader with several key foundationalcomponents. First, we introduce the historical debate between quantitative and qualitative researchers and how mixedmethods evolved into a legitimate research paradigm. Second, we present some of the basic characteristics of quantitativeand qualitative research and how they can be used in a complementary fashion. Third, we show the reader how our mixedmethods strategies t within a larger typology of mixed methods designs. Doing so will hopefully stimulate more thinkingabout the use of mixed methods for studies not only dealing with causation but for other research objectives as well (e.g.,prediction, description, and exploration). Fourth, we provide a brief overview ofmixedmethods data analysis strategies thatwill aid in ascertaining causation. We then present our mixed methods approaches for causation along with specicillustrations that are based on real life examples of sport management research. Examples come from three major areas ofsportmanagementwhich includemarketing, organizational behavior, and nance. The article is concludedwith a discussionand future directions.</p><p>1. The paradigm wars and the emergence of mixed methods research</p><p>Throughout the latter part of 20th century, social and behavioral scientists engaged in a fervent debate over thesupremacy of quantitative versus qualitative research (e.g., House, 1994; Johnson &amp; Onwuegbuzie, 2004; Tashakkori &amp;Teddlie, 1998). Each side argued from two distinct paradigmatic perspectives (i.e., a researchers world view of how best toobtain knowledge) (Tashakkori &amp; Teddlie, 1998). Quantitative purists, held to a positivist paradigmwhichmakes a variety ofassumptions including the following: there is a single objective reality, cause and effect relationships can be known(especiallywhen a randomized experiment is conducted), time and context free general...</p></li></ul>

Recommended

View more >