Globalization, democratization and knowledge production

3.7.1 Generalizability

Earlier in this paper, I noted that potential shortcomings of the case study method related to issues of validity. According to Cohen and Manion, a possible disadvantage of the case study method is that external validity is reduced and that it cannot be assumed that the results are applicable to other situations (1989, p. 129). However, I do not intend, nor presume, my study to be generalizable. The underlying assumption of a critical approach is that these particular cases were investigated in their own right. As Yin asserts, case studies are generalizable to theoretical propositions and not to populations or universes (1994, p. 10).

Notwithstanding this view, some degree of generalization is possible in the very sense that it is not. An interesting observation made by Smith (1995, p. 9) may be true of the institutions and participants in my study: “All schools are like no other schools, like some other schools, and like all other schools.” In that sense, the institutions and participants in my study may have some parallels with other higher education institutions in South Africa, and the rest of the world for that matter. Smith cites Spindler (1982) in support of this claim:

An in-depth study gives accurate knowledge of one setting not markedly dissimilar from other relevant settings, is likely to be generalizable in substantial degree to these other settings. (as cited in Smith, 1995 p. 304)
These universities do,in fact, represent three types of universities found in South Africa, historically white Afrikaans universities (HWU-A), historically white English universities (HWU-E) and historically black universities (HBU). Therefore, it is likely that an analysis of the selected cases may give us an idea of the constraints that researchers encounter in knowledge production at other South Africa universities, as well as of their response to globalization, the transformation policies and implementation processes, and of their visions for the future.

3.7.2 Triangulation

I contend that the triangulation of sites, methods and sources (participants) on several levels lent validity and credibility to this study. Data from the different sites have been compared and contrasted. Data from interviews and questionnaires have been contrasted and compared with one another, as well as with information from the document analysis and the ‘walking the campus’ observations. The triangulation of data from the different categories of participants (academics, graduate students, librarians, senior managers on the one hand, and different gender, race, language and age groups on the other) has also enhanced the trustworthiness of the data. As Banister et al. (1994) contend, triangulation facilitates richer and potentially more valid interpretations: “Exploration from a variety of sources using an appropriate combination of methods increases our confidence that it is not some peculiarity of source or method that has produced the findings” (p. 145).

3.7.3 Credibility

Careful attention to thoroughness, coherence and comprehensiveness (Schwandt, 1994, p. 122) in the research planning and procedures can confirm rigour and validity. By undertaking to present the findings to these universities, I have established what Guba and Lincoln (1985) refer to as “credibility,” a major criterion of trustworthiness (pp. 213, 219). In more conventional terms, this refers to the ‘internal validity’ of my research.

In qualitative research of this nature, the theory is emergent from the data and not vice-versa as in the positivist approach (Cohen and Manion, 1989, p. 39; Walsh, 1993, p. 52). Thus, as a case study researcher I could meet the researchers and other participants on their own ground and, through the interviews, work on making the operative theories of their research practices explicit through my study (Walsh, 1993, p. 52).

3.8 DATA ANALYSIS

According to Lofland and Lofland (1995), the data consist of whatever is recorded in write-ups, transcriptions, photographs and videos. For this reason, the recording task is considered to be “a crucial aspect of the naturalistic analysis of social life, the critical linchpin of our attempts” (pp. 67, 69). On the other hand, Powney and Watts (1987) claim that the concept ‘database’ refers to the “entire recorded and unrecorded data,” such as impressions gathered, ethos, posture, gestures and context. In my study, the data refer mainly to the recorded data (p. 143). The transcripts from the interviews, together with the completed questionnaires, constitute the major part of this data, while the field notes from my ‘walking the campus’ observations serve as complementary data.

The data were analysed according to the domains that emanated from the research design and that emerged from a preliminary analysis of the dominant themes or patterns in the data. These domains included the following: prevailing research cultures, managerialism/ entrepreneurialism in response to globalization and new local policies; access to research, disparities between HBUs and HWUs and innovations to overcome constraints, publishing and the public value of research; transformation of governance structures; implementation of equity and redress policies; Mode 2 type of research and ‘socially relevant’ research in partnership with communities and visions for the future. Although it may appear that I have identified too many domains, Merriam (1988) evinces that qualitative research, by its very nature strives to understand how all the parts work together to form a whole (p. 16). Therefore, the case study in qualitative research concentrates on many, if not all, variables present in a single unit (Merriam, 1988, p. 7). Analysis is not simply descriptive, it is “the detailed examination of the database… a creative, constructive affair” (Powney & Watts, 1987, pp. 158, 160). Powney and Watts point out that analysis should be consistent with the underlying philosophy of the research (ibid.).

The field notes from the interviews and observations were reviewed as soon as possible after the visit to fill in the gaps to constitute full field notes. Research methodologists warn of the dangers of memory loss with the passage of time and advise that full notes be written almost immediately after the interview or observation (Powney & Watts, 1987, p. 125; Lofland & Lofland, 1995, p. 91). The recorded interviews were transcribed while the quantitative data were analysed using a software programme.

In analyzing the transcripts, I paid attention not only to the words, but also to who spoke the words and the context in which they were spoken. Triangulation of the data and sources, careful coding and analysis, and allowing the voices of the participants to ‘come alive’ in the interpretation, lent further rigour to my approach. The data from the interviews were first analysed against the field notes, and then the questionnaires. The data were organised onto grids so that patterns and trends across the different participants could be examined. The data were then analysed according to the different categories of participants within and across the institutions, for example, academics and librarians within and across the three institutions. Thereafter, the patterns and trends across the categories were analysed. In this way, I developed case study profiles of individual institutions and then of themes, issues and trends across the institutions.

Thereafter, the data were analysed by drawing on the theoretical frameworks underpinning this study, the document analysis and the literature review. I used comparative analysis strategies as outlined by Slaughter (2001) to compare and contrast the findings both within and across the cases. Slaughter has presented alternative theories to the dominant modernist theories. These include political economic globalization theories, political sociology theories of Mann, knowledge/power theories of Foucault, narrativity theories and feminist theories. Feminist theory, Slaughter contends, is important for comparative higher education because it urges us to consider, amongst other things, equal rights, salary equity and comparable worth (2001, p. 406). Feminist development theory allows us to see the multiple voices as different ways of learning and thinking, which can be understood if we adopt the viewpoint of the “other” (ibid.). Slaughter asserts that a useful approach to studying gender cross-nationally might be to study the connections between faculty and the private or public sectors because “the neoliberal state valorizes the private sector and enables private, profit-taking organizations to indulge their preferences for hiring men” (2001, p. 406).

Scholars, like Altbach (1997), Scott (1999), Slaughter (2001) and Tierney (2001), suggest that the new theories call for mixed methods, multiple sites, quantitative and qualitative data, and a variety of analytical techniques. Tierney (2001) emphasises that he does not seek the theoretical dominance of postmodernism: “Indeed, theoretical preeminence of any singular view of the world is antithetical to the lineaments of postmodernism” (p. 366). Postmodernism, he argues, allows for heterogeneity so that localized and regional interpretations of different facets of knowledge, rather than supranational definitions become the organizing frameworks for comparative higher education (2001, p. 367). The interpretations of the findings were done in the context of the theoretical framework.

3.9 PRESENTATION OF FINDINGS

One set of findings is presented in the form of case studies of each of the three universities in the study in part two of this dissertation. The aim is create an understanding of the individual institution as it relates to my study. A further set of related findings is presented under specific themes where the data have been compared and contrasted across the institutions in part three of the dissertation. I have tried as far as possible to avoid any overlap of data but this has not been entirely possible because the case study is a part of the comparative study across institutions. I have thus taken the liberty to decide which data belong where, in other words what data best represent the individuals institution as a case studies, or the comparison of the three institutions.

Except for chapter eight and nine where I rely on the survey data, most of the data presented in the other chapters emanate from the interviews. As noted, I cite the participants as widely as possible in order to give voice to their authentic experiences. I have given the participants pseudonyms to protect their confidentiality. I denote these quotations and citations as (I: pseudonym). Where I need to specify that the data emanated from the survey, I use (S: pseudonym). In part four, I consider the implications of these findings. I make recommendations for improving access to research resources, building research capacity and for enhancing the implementation of the new policies at the universities so that these universities may contribute to the new democratic social order through their research programmes and knowledge producing processes.

10. 
    REFLECTIONS ON METHODOLOGY
    

During the course of my research, I found that critical postmodernism did not go far enough in helping me to understand and think about solutions to issues raised in this study. These issues related to the equity policies being implemented at the universities under study, managerialism as a project of neoliberalism obfuscating the redress intents of the national democratizing policies, and Mode 2 relevant research involving partnerships with local communities and indigenous knowledge systems, was not being applied in appropriate and a respectful ways (see further discussion below).

Critical postmodernism is useful in deconstructing notions of power and privilege and instances of gate-keeping, whereas decolonizing methodologies are required in exposing and ‘de-structing’ sites at which these obstacles to transformation are located and embedded. I use the term ‘de-structing’ to convey a more tangible, active intention to go beyond ‘deconstructing’ in order to uncover and remove racialised sites within the technologies, systems, structures and apparatuses of the university. I see “de-structing” as a more active way of “deconstructing”, and as a way that leads to practical results rather than theoretical propositions for problematising a concept or construct.

CHAPTER FOUR

“BECOMING MORE LIKE A BUSINESS” –

THE UNIVERSITY OF PORT ELIZABETH