Analyzing High-Profile Panel Discussion on Global Health: An Exploration with MAXQDA

Abstract: This article explores the potential of open-access videos available on the Internet as a data source to identify key areas in a rather broad topic in an international context. In order to understand the latest trends relating to a global issue, the study focuses on "global health," which is a growing interdisciplinary concept across health studies, international management and public policy research. By focusing on a video of a high-profile panel discussion in particular, the study also exemplifies main features of MAXQDA to assess the dialogues made by different speakers in the video. In doing so, the article considers key aspects related to the use of the existing videos and the nature of panel discussion, and how a combination of online videos and CAQDAS can pave the way for further research.

Key words: high-profile panel discussion; online videos; interdisciplinary topics; global health; bilingual data; MAXQDA

Against the background of the growth of video-sharing websites and the innovation of other Internet technologies, the accessibility and quality of open-source videos have dramatically increased, which has created a growing body of "video analysis" focusing on videos downloaded from the Internet (KNOBLAUCH, 2012). A great deal of information is now accessible online, not only from a variety of events, such as conferences and debates, in which a specific issue is discussed by experts, but also from high-profile interviews with those whom we would not normally have direct contact for individual interview. These public sources of information can serve as interesting data when analyzed systematically. However, while it has long been recognized that videos can provide extensive descriptions and serve as valuable sources for qualitative research (ROSENSTEIN, 2002), this means of data collection is still relatively underutilized in social sciences (LUFF & HEATH, 2012; SCHNETTLER & RAAB, 2008). This slow development is particularly the case in the field of management studies, as methodologies employed in this area is somewhat biased towards quantitative techniques to achieve higher generalizability. [1]

As the pace of change in the business environment is accelerating, however, research problems can be connected to various issues in a wider theme (HURMERINTA-PELTOMÄKI & NUMMELA, 2006). Key elements in what is occurring may not be readily identifiable using quantitative methods, because emerging elements are still in the process of coming to the fore. Moreover, research on international business and management has increasingly become interdisciplinary and involved idiosyncrasy in a specific context on the one hand, and complexity of various international factors on the other hand, because of proceeding globalization. Given these factors, qualitative methods will be advantageous to understand new occurrences for which no relevant models are yet available (GRIFFITHS, 1996). However, qualitative research often entails constraints, such as limited access to key informants that are essential to grasp complexity in an expansive topic, and to provide up-to-date analysis. Online videos, especially panel discussions of key people in a given area, can be valuable sources to obtain key actors' different perspectives, and using computer assisted qualitative data analysis (CAQDAS) packages for this type of video data will significantly increase the efficiency of the analytical process. [2]

The purpose of this article is to explore the potential of open-access videos available on the Internet as a data source to analyze global issues. This is an attempt to identify an issue in a rather broad topic as a starting point of a more comprehensive research by using existing videos and CAQDAS. In order to understand the latest trends relating to a global issue, this study focuses on "global health," which is a growing interdisciplinary concept across health studies, international management and public policy research. Global health can be broadly defined as "consideration of the health needs of the people of the whole planet above the concerns of particular nations" (BROWN, CUETO & FEE, 2006, p.62). Although the term "global health" itself is not an entirely new invention, this term has come into wide popular usage, as various types of organizations across borders are increasingly getting involved to tackle global health challenges (COOVADIA & HADINGHAM, 2005). [3]

To examine a recent trend surrounding global health, this study utilizes a video of a high-profile panel discussion available on YouTube. In doing so, the study also outlines the process of analyzing video data using MAXQDA, one of the major programs in qualitative research but relatively less known compared to NVivo among business and management researchers. Most CAQDAS packages have significantly improved features for handling audio/video data for the past several years (see SILVER & PATASHNICK, 2011 for a comprehensive examination of audiovisual tools of different CAQDAS packages), and MAXQDA has made some major improvements in recent versions. The article demonstrates some basic features of MAXQDA to show how an emerging and rather broad topic, such as global health, can be explored by using a video of public discussion. [4]

The following section considers key aspects related to analyzing video data with CAQDAS. Section 3 turns to a consideration of a high-profile panel discussion, and gives an overview of the concept of global health. Section 4 describes the methods and data. Section 5 presents results using MAXQDA. Section 6 provides a conclusion. [5]

There are two broad categories of video: 1. those recorded by researchers for their specific research projects, and 2. those that have been recorded by others, rather than researchers (e.g., by the actors that are in the videos themselves, or by third parties such as event organizers or professional/non-professional producers), and then later uploaded to YouTube or other websites (KNOBLAUCH, 2012). The former have been widely produced and used in qualitative research, and researchers can account for the process of production of such audio-visual data (ibid.). The latter includes "existing" videos (WARREN, 2009), where the researcher's position is purely that of viewer. An existing video might have been produced for different purposes, or used as a dissemination tool by the owner/distributor of the video, so that speakers in the video are not aware of the researcher's intentions or expectations. This lack of control in the process of recording will, however, enable the researcher to analyze narratives in the video in a more objective manner. [6]

Since the use of hypermedia is becoming a popular means of dissemination (PINK, 2001), not only individuals but also many organizations are increasingly keen to distribute a variety of information—for instance, the CEO's message, special interviews, in-depth reflections of debates—in the form of videos, and make them available on the Internet. The wealth of visual information available for qualitative research is a significant advantage of existing videos (STEYAERT, MARTI & MICHELS, 2012), and provides researchers with the access to rich narratives made by prominent figures in a given field. Moreover, another advantage of using videos, especially those depicting high-profile discussions, is topicality. Recent occurrences or critical incidents in a specific field may be discussed among key public figures at a public event and broadcasted on news media. In fact, videos can offer unprecedented access to new domains and activities (LUFF & HEATH, 2012). Researchers can utilize these materials to capture on-going changes and the latest situations. [7]

CAQDAS packages are especially useful for storing different types of data, such as video files and text documents, in one place, and analyzing them systematically. Until recently, most of the CAQDAS packages were text-based, and their features to integrate audio-video data were rather limited compared to Transana, which is especially designed for the qualitative analysis of digital video and audio data (cf. DEMPSTER & WOODS, 2011; SPIERS, 2004). However, since the late 2000s in particular, major CAQDAS package providers, such as ATLAS.ti, MAXQDA, and NVivo, have put great effort into improving the audio/video features to keep up with the rapid growth of multimedia. Now, most CAQDAS packages enable us to manage visual data and integrate them into writing tools more fully and easily than ever. For those who are newly venturing into CAQDAS-based research, the first decision to be made will be which package should be used or purchased, but as many experts in CAQDAS have pointed out, there is no one best package (e.g., KUCKARTZ & SHARP, 2011; SCHÖNFELDER, 2011; WEITZMAN, 2000). Since the CAQDAS packages allow researchers to apply different analysis strategies (SAILLARD, 2011), which one is most suitable will largely depend on what the researcher wants or needs to do. It is therefore very important to compare the software options carefully in light of the researcher's needs and the purpose of the project (WEITZMAN, 2000). [8]

Thanks to the capabilities of CAQDAS, as well as the variety of online videos available, there are a number of different ways to use video data, not only in qualitative research, but also in quantitative research, which is also a useful approach that collects measurable data from a number of videos (ROSENSTEIN, 2002). As for MAXQDA, the audio-visual feature was added from Version 10 alongside other new features, such as importing PDF documents and image files. MAXQDA offers good interface across different types of data and motivates the user to interact with the data (SAILLARD, 2011). [9]

While MAXQDA, alongside other CAQDAS packages, is capable of handling a large data, it is important to be selective in video analysis (HEATH, HINDMARSH & LUFF, 2010), because even a short video of a single event can provide a considerable volume of information, such as gestures, postures, and directions of gaze (FELE, 2012; GOFFMAN, 1963; KNOBLAUCH & SCHNETTLER, 2012; MONDADA, 2012). Those non-verbal languages could influence the analysis, even though the prime focus is on the speakers' narratives. Being selective in video analysis enables the researcher to provide more in-depth analysis, with a clearer focus. For example, Airi ROVIO-JAHANSSON (2007) selected a single video of one particular meeting in his analysis. In a group meeting or a panel discussion, moreover, it is essential to examine how the logical flow and discussion points shift in the conversation, which requires the researcher's careful attention. This is particularly important when the topic is communicated in a complex and interdisciplinary way. [10]

High-profile discussions have some distinct features, although they share something in common with focus groups or group interviews. In a way, the setting of a high-profile discussion (or panel discussion, in general) is similar to that of a focus group and group interview, where some participants and one moderator (or one, or more than one, interviewer) are present. At the outset, however, the difference between focus groups and group interviews should be noted. While the moderator of a focus group facilitates participants to interact with each other, the interviewer of a group interview typically asks questions of participants in turn (WILKINSON, 2004). In a panel discussion, the nature of the conversation and interactions will largely hinge upon how the moderator leads the conversation. [11]

Although panelists are expected to "discuss" a certain topic, it is quite common that the moderator invites each speaker's opinion in an orderly manner, and this is particularly so in a public event and a formal setting. Moreover, the speaker often discusses a certain issue by him-/herself to present his/her own opinion, rather than developing a discussion with other panelists. In such a case, the nature of the panel discussion can be more like a group interview. When observing a public video of a high-profile panel discussion, therefore, researchers need to consider whether the form of the discussion will suit the needs of a given research; for instance, to explore informed opinions; assess conflicting views; or compare different attitudes expressed by different organizational actors. [12]

When a video of a high-profile panel discussion is analyzed, it is vital to know not only the discussion topic but also the profiles of the speakers. Participants in existing videos are most likely to have been selected not by the researcher him-/herself, but by someone else (usually the organizer of the event) based on their interest. Thus, it is essential for the researcher to know the background of the speakers to develop a better understanding of why a speaker thinks and speaks in a certain fashion. In addition, it is equally important to take into account the context of the video. Context refers to "the set of circumstances surrounding an event or situation that help in its interpretation" (McGAUGHEY, 2004, p.529). As Hubert KNOBLAUCH and Bernt SCHNETTLER (2012) argue, analyzing a video does not merely consist of "content" analysis of its dialogue, but rather systematic analysis of the "context" in which interactions among participants are embedded. A speaker's actions and responses are shaped by the given context, and a sequence of action can only be understood with reference to the context (HERITAGE, 1984). In fact, a public video normally provides additional information, such as the layout of the venue in which interactions take place and the seating arrangement of the speakers, which may influence the flow of discussion. [13]

When a public discussion is videotaped, it should be noted that the sequence of discussion might be outlined in advance in a written agenda. Thus, compared to video footage of a direct field interview or a conversation between strangers, which are likely to take place with a wide range of contingency (MONDADA, 2009), dynamic interactions can be limited in a well-organized event. Nevertheless, Arnulf DEPPERMANN, Reinhold SCHMITT and Lorenza MONDADA (2010) argue that interaction and its transition in a formal event are still influenced by an actual performance, and that participants often collectively allow some divergence to emerge. Moreover, sophisticated speakers often express their views by spontaneously relating/comparing with other participants' opinions. It is therefore vital to be attentive to interactions in a real-time context, since interactions will unfold from the beginning towards the end of a video (MONDADA, 2012). [14]

While online pubic videos of high-profile panel discussions can provide rich and up-to-date information for qualitative research, there are also some drawbacks relating to this type of video. The potential problems are associated with practical, analytic, methodological and ethical aspects (HEATH et al., 2010). [15]

First, videos of public events suffer a classic concern in relation to the observer effect. Although speakers in public videos may not be directly influenced by a researcher's specific research interest, they are instead influenced by a wider public audience and cameras in the venue, and thereby they may act accordingly, or so-called "reactivity" (LAURIER & PHILO, 2006; SCHNETTLER & RAAB, 2008). Given the recognizable position held by high-profile panelists, their actions observed in a video may not be genuine responses to a stimulus at the scene (BORGER & CIOFFI, 1970), or their comments could be based on a well-edited draft speech. [16]

Second, while a video can display continuous interactions, camera movements, such as panning, tilting and zooming, are decided by the cameraman (MONDADA, 2012), so that when a public video is used as a data source, researchers cannot observe all actions and non-verbal language made by each speaker. This will be a main drawback compared to the researcher's own video-recoding, in which the researcher can take notes to augment the analysis of the video data in the process of production (SCHNETTLER & RAAB, 2008). [17]

Another drawback of using public videos is a concern of ethical issues, as many researchers already pointed out. This is particularly important when an image in a video is used as a visual presentation in a final output for publication, since the image might be copyrighted, and even thick description of video data could embarrass the speaker in the video as Gary ALBRECHT (1985) warns. Thus, selection of a suitable data source and what visual data are used to present the result will require great care (LUFF & HEATH, 2012). It is advisable to check the ownership of the image or possibly refer to the code of research ethics of researchers' institutions¹⁾. [18]

To demonstrate an analysis of a high-profile panel discussion on a global issue using an online video, this study explores discussions about "global health," a term that is growing in popularity within health management research. According to BROWN et al. (2006), there has been a surge in articles using the term "global health" since the 1990s, and it has become a fashionable term (KOPLAN et al., 2009). Despite its popularity, however, there is no standard definition of global health (JACK, GRIM & AULD, 2012). Richard SKOLNIK (2008, p.7) describes global health as "a global perspective on public health problems," but states that one does not have to worry about the definition, and that global health can be seen as part of public health. Conversely, others provide a clear distinction between "global health," "international health," and "public health" (cf. KOPLAN et al., 2009, pp.1993-1994): while "global health" looks at both domestic health issues and cross-border issues, "international health" focuses on health problems of foreign countries, especially low- and middle-income countries, but "public health" involves health issues within a particular country or community. However, the difference between "global health" and "international health" is somewhat vague, as the term "global health" is often used to refer to development assistance for improving health in low- and middle-income countries (e.g., RAVISHANKAR et al., 2009). [19]

In addition, it should be noted that the word "global" in global health is associated not only with location, but also with various organizations, such as governmental or international organizations, non-governmental organizations (NGOs), international foundations, and multinational corporations (BROWN et al., 2006; KOPLAN et al., 2009). The expansion of health matters, not only in terms of poverty-related underdeveloped health systems, but also communicable diseases such as HIV/AIDS, has increased the importance of different types of organizations in tackling the problems. Against this backdrop, healthcare organizations have to face institutional and technological challenges, while working with different stakeholders from diverse backgrounds (GORLI, KANEKLIN & SCARATTI, 2012). Importantly, the involvement of a large number of organizations has entailed a major change in the institutional landscape of global health (RAVISHANKAR et al., 2009). As a result, the key issues and actors involved in global health are becoming increasingly complex. [20]

In order to explore the latest situation relating to global health, the following section examines a video entitled "TICAD V Official Side Event: H igh-Level Panel 'From Okinawa to Tomorrow'". The panel discussion in this video clip was organized as a side event of the "Tokyo International Conference on African Development" (TICAD V), held in Japan in June 2013, and the video was subsequently published on YouTube by JCIEGlobalStudio as an official video from the event. This video was selected for this study from approximately 60 video clips that I collected for a broader research agenda for public-private interactions on global issues. It provides an ideal setting for the analysis of a wide-ranging discussion since the panel consists of experts in global health, but of different backgrounds. Therefore, the content of the discussion is very extensive reflecting each speaker's expertise. The panel discussion involves six speakers: a moderator and five panelists. They are affiliated with different global-health organizations. Table 1 outlines the profiles of the speakers.

Table 1: The profile of speakers in the high-profile debate on global health [21]

In the video, the moderator and the five panelists sat in a loose U-shaped layout on the stage, and the moderator was positioned on the right end, as the audience faces the stage. Since the moderator asked each panelist a question in turn, the form of the panel discussion was more similar to a group interview than a debate. The moderator asked a different but related question, while effectively connecting the question to each panelist's expertise and background. It is therefore interesting to see how the panelists' different views were gradually revealed as they spoke. [22]

First, the video was downloaded from YouTube using a free downloader, ClipGrab. The full-length video is approximately an hour and a half (including an opening remark, keynote speeches and a closing remark of the event), and the panel discussion part is about 43 minutes. Since this study focuses only on the panel discussion, the video was shortened by cutting other parts with Avidemux, a video-editing program. The short version was then imported to MAXQDA (Version 11). While the process of cutting a video is fairly easy, it should be noted that different CAQDAS packages may have different formatting requirements for video files (see SILVER & PATASHNICK, 2011, for more details). Regarding the data size, there used to be some limitation in importable data size (cf. SCHÖNFELDER, 2011). However, functionality appears to have been improved in a newer version. The shortened video used in this study is still larger than 560MB, but there was no problem with importing and operating it within the MAXQDA environment. Nevertheless, if a very long video or a significantly large number of short videos are used, it will be better to check whether the data is efficiently handled in a given CAQDAS package. If the total of the final data size is uncertain at the beginning of the project, an alternative approach is to store large files/links externally. [23]

In an international panel discussion, speakers may speak different languages. In the video used in this study, English and Japanese were spoken. Although simultaneous interpretation services were provided to the panelists and the audience at the venue, only the original voices are available in the video. An interesting incident in this video is that Speaker E switched languages, from English to Japanese (his native language), at the end of his speech as he addressed the political leaders in the venue, and stressed the importance of providing support to global health. Language switching often occurs when speakers emphasize a point, or highlight professional and culturally specific topics (WELCH & PIEKKARI, 2006). [24]

In MAXQDA, there can be several options for transcribing different languages, for instance: 1. transcribe and code in the original language, and translate only the results of the analysis into English (given the language to be used in the final output is English); 2. transcribe in the original language, and then provide an English translation immediately after it on the transcript document; or 3. translate foreign languages into English at first hand, while transcribing the original language in a memo (of which function will be descried below). The first option tallies with Barbara CZARNIAWSKA's (2004) translation strategy and this appears to be most advisable, especially if the speech is related to subtleties of cultural norms and values, and rough translation could negatively affect analysis. [25]

However, which way would suit most depends on the frequency of language switching and the length of talk in a different language. If language switching occurs only once or very few times in a video, the second option, i.e., transcribing in the original language (Japanese in the case of this video), and then providing a literal translation in English on the transcript can be useful to grasp what the speaker said at a glance, as shown in Figure 1.

Figure 1: Transcribing in MAXQDA: document browser and memo boxes. Please click here for an increased version of Figure 1. [26]

If there are frequent switches of language, or if more than two languages are used, a verbatim translation at each occurrence of language switching would be time consuming and make the coding process confusing. In such a case, the third option may be used. Yet, when the original foreign language is translated, it requires greater attention to the nuances of the language in order to preserve the meanings of the words originally spoken (HARZING et al., 2005; XIAN, 2008). [27]

The "memo" system mentioned above is a very useful tool in MAXQDA. In fact, memo functionality can be seen as one of the key advantages that MAXQDA provides (LEWINS & SILVER, 2007; SCHÖNFELDER, 2011). It works in the same way as a post-it note, on which one can write anything—ideas, definitions, comments, external links etc.—in memo form, and attach them to any documents, text segments, or codes. In addition, MAXQDA's "memo manager" provides highly effective control over assigned memos, and includes a lexical search function across memos. Memos are particularly useful in transcribing video data when it comes to taking notes regarding nonverbal communication and other additional information, and attaching them to the exact place in the transcript where they are observed. In Figure 1 above, the small post-it note icons are shown on the y-axis on the left-hand side of the document browser, and one of them is opened in a small content-preview window in the middle. [28]

Because of memo's flexibility and great capacity (each memo can contain up to 32 pages of text), researchers may be tempted to describe whatever interesting in the source video. Nevertheless, it is important not to over-describe every detail. Video data contain plenty of verbal, nonverbal and contextual information, and speakers in a discussion on a weighty topic may comment on various factors, including less important elements for the researcher's analytic focus. Hence, transcribing every small detail could result in mislaying key points. As HEATH et al. (2010) suggest, transcription needs to provide a basis for developing observation, rather than simply representing too many aspects of the activity within the data source. Elinor OCHS (1979) also maintains that transcription is a selective process, which should reflect theoretical goals. Hence, it is essential to ensure a clear filtering process, because an overly detailed transcript can be difficult to analyze. [29]

The "timestamp" tool is another essential tool in transcribing video data. It enables us to play back and forth, or jump to a specific part of the video easily. Timestamps are commonly used in audio-video files to mark time positions when different speakers begin to speak. In a panel discussion, one speaker may give a lengthy explanation, so that timestamps can be created more often, possibly wherever a notable comment is made, within a speaker's dialogue. This makes revisiting the exact point of the comment easier. The effective use of timestamps also provides a clearer picture of the logical transition during the discussion; some speakers may explain facts chronologically, while others may start with a broad background and then narrow down their focus to more specific issues, or vice versa. Observing the logical structure of each speaker's speech is important to grasp each speaker's approach to the issue. [30]

This section explores the issues in global health using the video of the high-profile discussion described above, while exemplifying the process of data analysis by using MAXQDA. It should be noted that MAXQDA, as well as any CAQDAS packages, is not a method but a tool to assist an analytical method (GIBBS, 2004). There are, of course, many approaches to video data depending on the research question, but rather than detailing various technical aspects, this section aims to identify key issues in the recent dialogue on global health by demonstrating four main features of MAXQDA—i.e., codeline, code relations, code-subcode-segment model, and code theory model. [31]

As many qualitative researchers are well aware, coding in a qualitative study tends to be a long process. Sharon BONG (2002) warns that there is a pitfall that researchers might regard coding as an end, rather than a means of theory building. As Raymond LEE and Nigel FIELDING (1996) argue, the prime aim of coding is data reduction. Therefore, simply indexing or coding line by line does not necessarily achieve conceptualization (SAILLARD, 2011). In analyzing a video of a panel discussion on a complex topic in particular, it is crucial to understand what each speaker is trying to say, and grasp a key point of the argument. In this study, MAXQDA's codeline browser was operated after careful coding in order to obtain a comprehensive view of the codes assigned to the transcription. The codeline browser visualizes the structure of a document and overlaps of codes in different paragraphs. This feature is very useful throughout the analytical process of a panel discussion in terms of assessing the flow of conversation as well as each speaker's discussion points. It illustrates how the discussion was organized, how the focus of the discussion shifted, and what codes were associated with each speaker, in a sequential manner. For the transcription of the panel discussion, this was done by coding each speaker's part with a relevant name, such as "Moderator," "Speaker A," "Speaker B," and so on, in the document. Figure 2 shows the code line for the selected codes assigned to the document.

Figure 2: A panel discussion at a glance by codeline browser. Please click here for an increased version of Figure 2. [32]

In Figure 2, the first six codes in the code system show each speaker (moderator, and Speakers A to E). The colored cells on the x-axis indicate that the moderator and the panelists spoke in turns; the moderator (assigned a black color vis-à-vis grey for the other speakers' cells) spoke first (under Paragraph 1), followed by Speaker A, and back to the moderator, then Speaker B, and so forth. Paragraph 11 shows the moderator's last question, which was addressed to two speakers, i.e., Speakers A and B, who responded in a successive way, as can be seen in Paragraphs 12 and 13. Paragraph 14 is the moderator's brief closing message. Therefore, no code was assigned to the y-axis of this paragraph. [33]

The moderator initiated the panel discussion by mentioning financial issues, because, retrospectively, the biggest problem in the field of global health during the 1990s was a lack of financial resources to tackle the multiple plagues in poor countries (GARRETT, 2007). However, a large number of international funds have been established since then, and more resources are now being provided to global health (COOVADIA & HADINGHAM, 2005; RAVISHANKAR et al., 2009). As a result, the key issues in global health have been shifting to other areas, including a wider range of priorities for different parties (SRIDHAR, 2012). This trend is reflected in the panel discussion. As the y-axis of Paragraphs 1 to 13 exhibits, each speaker commented on a variety of topics. For example, Speaker C referred to various policies and different communities, whereas Speaker D explained technology-related matters. The next step of this analysis will narrow down the focus by accessing the data in more detail. [34]

Now we will look into what the speakers talked about in relation to several issues in global health, and how often they referred to the topic/issue during the discussion; in other words, the associations between each speaker and different codes, as well as the frequency of the codes in the document. In doing so, the code relations browser is used. It is one of MAXQDA's visual tools, which indicates the frequency of the overlaps by the size of the square/circular nodes in a chart; it can also show the frequency numerically. The latter visualization is shown in Figure 3.

In Figure 3, the speakers are listed on the x-axis, while the four main codes, namely "investment," "technology," "policy issues" and "communities" and their sub-codes, are included on the y-axis. Sub-codes can be hidden/shown by clicking a plus/minus button at the top of the code. In Figure 3, only the sub-codes for "communities" are shown to demonstrate this. The value created by the code relations browser can be exported to an Excel file to obtain the total number of codes, as Table 2 shows.

This indicates how different codes are associated with each speaker. The code relations clearly reflect each speaker's professional background. For example, Speaker C, who was described by the moderator as "political animal" in the video, did not mention technology, while Speaker D, who comes from the vaccines business background, did not comment on communities. Another interesting feature we can observe here is that Speaker A covered various issues ranging all the four main codes, which appears to be more informative compared to Speaker E, for instance. One reason for the disparity in the number of codes between Speaker A and E is that Speaker A spoke much faster than Speaker E in a given time; another reason is Speaker E's speech spent more time on detailing other issues rather than the most coded topics, such as the different types of diseases and their current situations (as he is a medical doctor). In a way, the perspective of the generalist (Speaker A) versus that of the specialist (Speaker E) can be an interesting aspect to taken into consideration in analyzing a panel discussion. [37]

As the code relations show, overall the code group "policy issues" took up the biggest proportion of codes than the other main codes. Since the code category of "policy issues" is rather broad, it is now necessary to examine each sub-code in this category in more detail through the code-subcode-segment model. This model is one of the free-mapping tools, which are called MAXMaps in MAXQDA. The model investigates instances of sub-codes within a main code. In its standard format, the model displays the main code and its sub-codes in a radial symmetry connected by simple straight lines, but the diagram can be re-arranged by adding text, changing lines and moving objects. Less important sub-codes can be hidden in a different layer, or simply deleted. In Figure 4, six sub-codes in "policy issues" are activated. Each sub-code is connected to coded segments (shown as small, shield-shaped objects). The content of a segment can be viewed by hovering over a shield, so that it is easy to grasp what was actually said by a speaker in the exact place where the code is assigned.

Figure 4: Mapping "policy issues" in code-subcode-segment model. Please click here for an increased version of Figure 4. [38]

Regarding the actions to be taken for advancing global health, most speakers reaffirmed the importance of both national health systems and externally funded health programs, although, in the past, internal and external courses of action for global health policy were often considered separately. Reflecting the tendency that global health policy has been "fragmented and verticalized" in recent years (OLLILA, 2005), some panelists also touched upon various aspects of the "holistic approach," albeit rather briefly. A notable sub-code in the policy issues is "leadership," which is the most coded sub-code in this main code, and was mentioned by all the speakers. (This can also be seen in Figure 2.) [39]

Since the code-subcode-segment model displays a maximum of ten segments (which can be prioritized either by weight or segment size), one may want to check other segments to verify the importance of the selected segments. This can be done by using a lexical search function in MAXQDA. There are 72 segments containing the words "leader/leaders" or "leadership" in this 43-minute video file. With the help of the lexical search result, which provides a preview of each segment and its position in the document, some segments that are unrelated to the speakers' arguments (such as complement greetings to the host government of the conference) are excluded. Even after deducting these less important segments, the frequency in the data is still noteworthy, especially when compared to the previous World Health Report (WHO, 2010), which mentions "leadership" only twice in the whole volume of over 120 pages. [40]

The last step in this study is to create a conceptual map. Based on the analysis so far, we can hypothesize that "leadership" may be one of key issues emerging in global health. While leadership is not a new topic in business and management studies, this term has tended to be used in a loose manner in heath studies, often without a clear definition. Notably, a consensus was made at a conference, which defines leadership as "providing strategic direction and opportunities for participation in developing healthy public policy, mobilizing and managing resources for health promotion, and building capacity" (ALLEGRANTE et al., 2009, p.479). More recently, some experts have also argued that health goals in the Millennium Development Goals (MDGs)²⁾ have not been adequately achieved because of a lack of effective global leadership (RICHARD et al., 2011). Given the largely uncoordinated resource allocation in global health (FRANCO-PAREDES, ZEULI, HERNÁNDEZ-RAMOS & SANTOS-PRECIADO, 2010), the prime focus of leadership should be on how to set priorities to promote global health (SRIDHAR, 2012). [41]

Grounded on the results obtained through MAXQDA, a conceptual map can be drawn using MAXQDA's code theory model. There are two ways to execute this: 1. drag codes onto the map one by one, or 2. drag a main code and its sub-codes onto the map together. In this project file, all of the main codes mentioned above are organized under the first-level code with the broad code name "key issues in global health." Thus, this parent code, together with all of the attached memos, was selected for insertion into the map window, and then they were reorganized on the map browser, as Figure 5 shows. Figure 5 still shows memos for reference purposes, but these can be hidden or deleted for clearer visual presentation.

Figure 5: Conceptual map. Please click here for an increased version of Figure 5. [42]

The conceptual model may locate investment, technology, and communities as the main areas that leaders should address. In addition to this priority-setting, one might ask "who should/can take the leadership role?" In fact, given the large milieu of global health, there can be many entities that would assume a leadership role. Traditionally, UN systems, especially WHO, have promoted the role of leadership in global public health (YACH & BETTCHER, 1998), but the role of UN systems as a whole has been declining (LANCET, 2009; RUCKERT & LABONTÉ, 2012). In the meantime, the importance of other organizations in the private sectors and in NGOs is increasing (HEIN, BARTSCH & KOHLMORGEN, 2007; LURIE, 2012; OLLILA, 2005). Further analysis of global health can focus on specific organizations and deepen the analysis by applying this conceptual model to assess the extent to which key organizations are capable of undertaking the leadership in the respective areas. [43]

This article aimed to explore the usefulness of online videos, especially those of high-profile panel discussions in an interdisciplinary subject, to reveal key issues of a current global issue by using CAQDAS. In doing so, MAXQDA was employed to analyze the recent dialogue about global health. Given the broad nature of global heath and the various factors and diverse actors related to it, a systematic approach to data by using a CAQDAS package is effective not only in terms of identifying emerging issues, but also visualizing results. Drawing on the results, the study found that leadership appears to be key to the further development of global health. [44]

Of course, the article is not comprehensive. Instead, it precised general aspects related to the use of public videos of panel discussions as a data source, and demonstrated selected features of MAXQDA to assess the dialogues made by the speakers with different expertise. Since CAQDAS packages provide researchers with great flexibility, there are certainly other approaches and coding schemes available. [45]

One limitation, which this article did not detail but is still important, relates to the potentially different notions that speakers may have on a topic. When panelists from different backgrounds discuss a global issue in a temporary (and notably international) setting in the form of panel discussion, they may be drawing on different definitions or even conflicting perceptions of the same thing or a certain piece of terminology. In the video used in this study, for example, Speakers A and B expressed quite different views—a positive and a negative view respectively—about the credibility of political leaders. This means that "political leaders" can be either a positive (e.g., reliable, powerful) or negative (unreliable, corrupted) connotation. Even though it may not be explicitly expressed, their tone of voice and anecdotes they used in the discussion reflect their views. In this regard, it is essential to note that participants' responses only make sense against the person's lived experience, and his/her society at the time (GRIFFITHS, 1996; HARBOUR & KISFALVI, 2012). In order to investigate each speaker's narrative in more depth, text retrieval, rather than code retrieval, may be more useful (cf., COFFEY & ATKINSON, 1996). All of the major CAQDAS packages provide sophisticated text retrieval functions, so that the use of such a tool may reveal interesting findings from this type of video data. Furthermore, while a selective attitude towards video data is important, using multiple videos of different discussions among different speakers on a specific topic is indeed an effective alternative approach to achieve a more comprehensive analysis. [46]

Against the background of the increasing importance of a multi-method approach as a result of the blurring boundaries between disciplines (HURMERINTA-PELTOMÄKI & NUMMELA, 2006), using online videos of high-profile panels can offer an interesting research setting in which a variety of issues are discussed among well-informed people. Although this does not mean that existing videos could replace first-hand interviews, the analysis of public videos—if the source is properly selected—can serve as an effective means by which to grasp a hot topic, and pave the way for further investigation of it (STEYAERTet al., 2012). Given an extensive range of information available on the Internet, a combination of online videos and CAQDAS can significantly enrich qualitative research. [47]

1) For this study, for instance, I contacted the Research Ethics Office of my university to check the policy for image-based research, and followed an advice that images captured from the video should not be used unless they are absolutely necessary for an analysis of this article. <back>

2) The Millennium Development Goals (MDGs) are international development goals established by the United Nations in 2000, which encompass eight sections of major development issues such as; poverty reduction, education, gender, child mortality, maternal health, HIV/AIDS and malaria, sustainability, and global partnership. More details can be found at http://www.un.org/millenniumgoals/. <back>

Albrecht, Gary L. (1985). Videotape safaris: Entering the field with a camera. Qualitative Sociology, 8(4), 325-344.

Allegrante, John P.; Barry, Margaret M.; Airhihenbuwa, Collins O.; Auld, M. Elaine; Collins, Janet L.; Lamarre, Marie-Claude; Magnusson, Gudjon; McQueen, David V.; Mittelmark, Maurice B. & Galway Consensus Conference (2009). Domains of core competency, standards, and quality assurance for building global capacity in health promotion: The Galway consensus conference statement. Health Education & Behavior, 36(3), 476-482.

Bong, Sharon A. (2002). Debunking myths in qualitative data analysis. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 3(2), Art. 10, http://nbn-resolving.de/urn:nbn:de:0114-fqs0202107 [Date of access: November 14, 2013].

Borger, Robert & Cioffi, Frank (Eds.) (1970). Explanation in the behavioural sciences. London: Cambridge University Press.

Brown, Theodore M.; Cueto, Marcos & Fee, Elizabeth (2006). The World Health Organization and the transition from "international" to "global" public health. American Journal of Public Health, 96(1), 62-72.

Coffey, Amanda & Atkinson, Paul (1996). Making sense of qualitative data: Complementary research strategies. Thousand Oaks, CA: Sage.

Coovadia, Hoosen M. & Hadingham, Jacqui (2005). HIV/AIDS: Global trends, global funds and delivery bottlenecks. Globalization and Health, 1(13), http://www.globalizationandhealth.com/content/1/1/13 [Date of access: October 12, 2013].

Czarniawska, Barbara (2004). Narratives in social science research. London: Sage.

Deppermann, Arnulf; Schmitt, Reinhold & Mondada, Lorenza (2010). Agenda and emergence: Contingent and planned activities in a meeting. Journal of Pragmatics, 42, 1700-1718.

Dempster, Paul G. & Woods, David K. (2011). The economic crisis though the eyes of Transana. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 12(1), Art. 16, http://nbn-resolving.de/urn:nbn:de:0114-fqs1101169 [Date of access: November 22, 2014].

Fele, Giolo (2012). The use of video to document tacit participation in an emergency operations centre. Qualitative Research, 12(3), 280-303.

Franco-Paredes, Carlos; Zeuli, Julia; Hernández-Ramos, Isabel & Santos-Preciado, Jose I. (2010). Preserving idealism in global health promotion. Global Health Promotion, 17(4), 57-60.

Garrett, Laurie (2007). The challenge of global health. Foreign Affairs, 86(1), 14-38.

Gibbs, Graham R. (2004). CAQDAS (Computer-Assisted Qualitative Data Analysis Software). In Michael S. Lewis-Beck; Alan Bryman & Tim Futing Liao (Eds.), The SAGE encyclopedia of social science research methods (pp.88-90). Thousand Oaks, CA: Sage.

Goffman, Erving (1963). Behavior in public places: Notes on the social organization of gatherings. New York: Free Press.

Gorli, Mara; Kaneklin, Cesare & Scaratti, Giuseppe (2012). A multi-method approach for looking inside healthcare practices. Qualitative Research in Organizations and Management, 7(3), 290-307.

Griffiths, Frances (1996). Qualitative research: The research questions it can help answer, the methods it uses, the assumptions behind the research questions and what influences the direction of research. A summary of the panel discussion at the conference "Exploring qualitative research in general practice". Family Practice, 13(Suppl. 1), S27-S30.

Harbour, Michelle & Kisfalvi, Veronika (2012). Looking desperately for courage or how to study a polysemic concept. Qualitative Research in Organizations and Management: An International Journal, 7(2), 166-188.

Harzing, Anne-Wil with 32 country collaborators (2005). Does the use of English-language questionnaires in cross-national research obscure national differences? International Journal of Cross Cultural Management, 5(2), 213-224.

Heath, Christian; Hindmarsh, John & Luff, Paul (2010). Video in qualitative research: Analysing social Interaction in everyday life. Los Angeles, CA: Sage.

Hein, Wolfgang; Bartsch, Sonja & Kohlmorgen, Lars (Eds.) (2007). Global health governance and the fight against HIV/AIDS. Basingstoke: Palgrave Macmillan.

Hurmerinta-Peltomäki, Leila & Nummela, Niina (2006). Mixed methods in international business research: A value-added perspective. Management International Review, 46(4), 439-459.

Jack, Leonard, Jr.; Grim, Melissa & Auld, M. Elaine (2012). Health promotion practice expands focus on global health promotion. Health Promotion Practice, 13(3), 289-292.

JCIEGlobalStudio (2013). TICAD V official side event: High-level panel "From Okinawa to Tomorrow", http://www.youtube.com/watch?v=Ul7OCtqwFrk [Date of access: July 10, 2013].

Knoblauch, Hubert (2012). Introduction to the special issue of qualitative research: Video-analysis and videography. Qualitative Research, 12(3), 251-254.

Knoblauch, Hubert & Schnettler, Bernt (2012). Videography: analysing video data as a "focused" ethnographic and hermeneutical exercise. Qualitative Research, 12(3), 334-356.

Koplan, Jeffrey P.; Bond, T. Christopher; Merson, Michael H.; Reddy, K. Srinath; Rodriguez, Mario Henry; Sewankambo, Nelson K. & Wasserheit, Judith N., for the Consortium of Universities for Global Health Executive Board (2009). Towards a common definition of global health. The Lancet, 373, 1993-1995.

Kuckartz, Anne M. & Sharp, Michael J. (2011). Responsibility: A key category for understanding the discourse on the financial crisis—analyzing the KWALON data set with MAXQDA 10. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 12(1), Art. 22, http://nbn-resolving.de/urn:nbn:de:0114-fqs1101222 [Date of access: October 12, 2013].

Laurier, Eric & Philo, Chris (2006). Natural problems of naturalistic video data. In Hubert Knoblauch, Bernt Schnettler, Jürgen Raab & Hans-Georg Soeffner (Eds.), Video analysis: Methodology and methods—Qualitative audiovisual data analysis in sociology (pp.183-192). Frankfurt/M.: Peter Lang.

Lee, Raymond M. & Fielding, Nigel (1996). Qualitative data analysis: Representations of a technology: A comment on Coffey, Holbrook and Atkinson. Sociological Research Online, 1(4), http://www.socresonline.org.uk/1/4/lf.html [Date of access: November 19, 2013].

Lewins Ann & Silver, Christina (2007). Using software in qualitative research: A step-by-step guide. Los Angeles, CA: Sage.

Luff, Paul & Heath, Christian (2012). Some "technical challenges" of video analysis: Social actions, objects, material realities and the problems of perspective. Qualitative Research, 12(3), 255-279.

Lurie, Sue G. (2012). Global health equity and advocacy: The roles of international non-governmental organizations. Health, Culture and Society, 2(1), 104-114.

McGaughey, Sara L. (2004). Writing it up: The challenges of representation in qualitative research. In Rebecca Marschan-Piekkari & Catherine Welch (Eds.), Handbook of qualitative research methods for international business (pp.529-550). Cheltenham: Edward Elgar.

Mondada, Lorenza (2009). Emergent focused interactions in public places: A systematic analysis of the multimodal achievement of a common interactional space. Journal of Pragmatics, 41(10), 1977-1997.

Mondada, Lorenza (2012). Video analysis and the temporality of inscriptions within social interaction: The case of architects at work. Qualitative Research, 12(3), 304-333.

Ochs, Elinor (1979). Transcription as theory. In Elinor Ochs & Bambi B. Schieffelin (Eds.), Developmental pragmatics (pp.43-72). New York: Academic Press.

Pink, Sarah (2001). More visualising, more methodologies: On video, reflexivity and qualitative research. Sociological Review, 49(4), 586-599.

Ravishankar, Nirmala; Gubbins, Paul; Cooley, Rebecca J.; Leach-Kemon, Katherine; Michaud, Catherine M.; Jamison, Dean T. & Murray, Christopher J.L. (2009). Financing of global health: Tracking development assistance for health from 1990 to 2007. The Lancet, 373, 2113-2124.

Richard, Fabienne; Hercot, David; Ouédraogo, Charlemagne; Delvaux, Thérèse; Samaké, Salif, van Olmen, Josefien; Conombo, Ghislaine; Hammonds, Rachel & Vandemoortelei, Jan (2011). Sub-Saharan Africa and the health MDGs: The need to move beyond the "quick impact" model. Reproductive Health Matters, 19(38), 42-55.

Rovio-Johansson, Airi (2007). Post-acquisition integration: Ways of sensemaking in a management team meeting. Qualitative Research in Organizations and Management: An International Journal, 2(1), 4-22.

Ruckert, Arne & Labonté, Ronald (2012). The financial crisis and global health: The International Monetary Fund's (IMF) policy response. Health Promotion International, 28(3), 357-366.

Saillard, Elif K. (2011). Systematic versus interpretive analysis with two CAQDAS packages: NVivo and MAXQDA. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 12(1), Art. 34, http://nbn-resolving.de/urn:nbn:de:0114-fqs1101345 [Date of access: October 12, 2013].

Schnettler, Bernt & Raab, Jürgen (2008). Interpretative visual analysis: Developments, state of the art and pending problems. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 9(3), Art. 31, http://nbn-resolving.de/urn:nbn:de:0114-fqs0803314 [Date of access: October 12, 2013].

Schönfelder, Walter (2011). CAQDAS and qualitative syllogism logic: NVivo 8 and MAXQDA 10 compared. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 12(1), Art. 21, http://nbn-resolving.de/urn:nbn:de:0114-fqs1101218 [Date of access: October 12, 2013].

Silver, Christina & Patashnick, Jennifer (2011). Finding fidelity: Advancing audiovisual analysis using software. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 12(1), Art. 37, http://nbn-resolving.de/urn:nbn:de:0114-fqs1101372 [Date of access: October 12, 2013].

Skolnik, Richard L. (2008). Global health 101 (Essential public health). Burlington, MA: Jones & Bartlett Learning.

Steyaert, Chris; Marti, Laurent & Michels, Christoph (2012). Multiplicity and reflexivity in organizational research: Towards a performative approach to the visual. Qualitative Research in Organizations and Management: An International Journal, 7(1), 34-53.

Warren, Samantha (2009). Visual methods in organizational research. In David A. Buchanan & Alan Bryman (Eds.), The Sage handbook of organizational research methods (pp.566-582). London: Sage.

Weitzman, Eben A. (2000). Software and qualitative research. In Norman K. Denzin & Yvonna S. Lincoln (Eds.), Handbook of qualitative research (pp.803-820). Thousand Oaks, CA: Sage.

Welch, Catherine & Piekkari, Rebecca (2006). Crossing language boundaries: Qualitative interviewing in international business. Management International Review, 46(4), 417-437.

Wilkinson, Sue (2004). Focus group research. In David Silverman (Ed.), Qualitative research: Theory, method and practice (2nd ed., pp.177-199). London: Sage.

World Health Organisation (WHO) (2010). The World health report, 2010: Health systems financing—the path to universal coverage. Geneva: WHO.

Xian, Huiping (2008). Lost in translation? Language, culture and the roles of translator in cross-cultural management research. Qualitative Research in Organizations and Management: An International Journal, 3(3), 231-245.

Yach, Derek & Bettcher, Douglas (1998). The globalization of public health, II: The convergence of self-interest and altruism. American Journal of Public Health, 88(5), 738-741.

Dr. Faith HATANI is lecturer in international business at Manchester Business School, the University of Manchester, UK. Her research focuses on the role of international business in economic development. She is interested in qualitative analysis of public-private interactions in achieving development goals, and foreign direct investment in infrastructure, in particular in developing countries.

Manchester Business School
University of Manchester
Booth Street West, M15 6PB, UK

Hatani, Faith (2015). Analyzing High-Profile Panel Discussion on Global Health: An Exploration with MAXQDA [47 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 16(1), Art. 14,
http://nbn-resolving.de/urn:nbn:de:0114-fqs1501148.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License

	Moderator	Speaker A	Speaker B	Speaker C	Speaker D	Speaker E
Position and organization	Senior Fellow for Global Health, Council on Foreign Relations	Executive Director, Global Fund	Executive Director, Roll Back Malaria Partnership	Deputy Executive Director UNAIDS	President and CEO, International AIDS Vaccine Initiative	Regional Director Emeritus, World Health Organization for the Western Pacific
Type of organization	Non-profit think tank	International financing organization	Global network (focusing on Africa)	International organization	Public-private partnership	International organization