Paradigms
4.4 Research design
165
the same study”. To avoid such a scenario, the present study used mixed methods triangulation design procedures, using a multilevel research variant where the qualitative and quantitative methods were used to address different levels within the case studies (Tashakkori and Teddlie 1998:48). As preempted in the previous chapter, a considerable number of scholars in records and archives management have used the MMR to address related research problems successfully such as Kabata 2019;
Musembe 2019; Marutha 2016; Maseh 2015; Laughton 2011; Kalusopa 2011; Garaba 2010 and Luyombya 2010. On the flipside however, mixed methods research is relatively new and clouded in controversy as this is an evolving methodology. The development of really integrated qualitative/quantitative methods remains a problem to be solved and universally accepted suggestions for how to integrate both approaches in one method are still awaited (Flick 2018:74).
166
(Creswell and Clark 2017:184-195). In view of these, researchers should assess their proposed research design to determine its suitability before going into the field for data collection. The present study adopted the multiple case study research design.
4.4.1 Case study design
According to Rose, Spinks and Canhoto (2014:102), the word ‘case’ in research means ‘an instance of’. Therefore, case study research design can be described as an investigation of one or more ‘instances of’ something that constitutes the case(s) in a given study. A broader definition is provided by Creswell (2013:97) who defines case study method as an exploration of “a real-life, contemporary bounded system (a case) or multiple bounded systems (cases) over time, through detailed in-depth data collection involving multiple sources of information”. Bryman (2012:66) defined case studies as methodological approaches involving intensive, detailed and in-depth exploration of specific ‘bounded’ systems, using numerous data collection procedures to gather information systematically on how the systems function or operate. In retrospect, a case can be as simple as an individual or group, or as complex as an organization, neighborhood or culture, or it may be something more abstract like a programme, an event or activity (Rose, Spinks and Canhoto 2014:102). Focusing on the context or site of a given practice is particularly applicable to practice theory (Miles 2015:311), which in the context of this research entails digital archiving practices in universities. Thus, case studies provide contextual knowledge and participants’ versions of practice that are drawn together from their interactions, actions, voices and practices in the study site (Miles 2015:311).
Yin (2009:46) categorized case studies as exploratory, explanatory or descriptive in nature and distinguished the following four types of case study designs:
1) Single case (holistic) designs – where a single unit of analysis is selected to represent a unique or critical case. One can also select a single case as a representative or typical case or one which has not been considered before;
2) Single-case (embedded) designs – this involves more than one unit of analysis within a single case. The sub-units have been found to add significant opportunities for extensive analysis, enhancing the insights into the single case;
3) Multiple-case (holistic) designs – where a study contains more than one case; and
167
4) Multiple-case (embedded) designs – involves several units of analysis within the multiple cases.
McKemmish and Gilliland (2013:94-5) opine that in-depth single case or multiple case studies are appropriate for exploring current recordkeeping practices. This study adopted a multiple case (embedded) design described as case study research, comprising several instrumental bounded cases, carefully selected to develop a more in-depth, better understanding of the phenomenon than can be provided by a single- case study (Encyclopedia of Case Study Research 2010:584). Data was collected from various units within six different universities namely: University of Nairobi (UoN) located along University Way in Nairobi central business district; Moi University (MU) located in Kesses, 35 kilometres from Eldoret town; Kenyatta University (KU) located in Kahawa, about 20 kilometres from Nairobi's city centre; Maseno University (MSU) based in Maseno, near Kisumu city; Jomo Kenyatta University of Agriculture and Technology (JKUAT) located in Juja, along the Nairobi-Thika highway, and;
Egerton University (EU) located in Njoro, near Nakuru town. The choice of a multiple-case research design was appropriate since it enabled the researcher to capture a more holistic and contextual view of digital archives management in the units under study. This particular design was also selected with lateral replication in mind, whereby the cases selected were similar and the predicted results were also similar. The data thus collected provided adequate guidance towards developing a framework for digital archiving applicable for use in public universities in Kenya.
The case study sites were purposively selected based on the criteria that they were among the 23 government accredited public universities in Kenya. They were also selected in the order of their years of establishment (that is, the six earliest universities) as follows: University of Nairobi (UoN) (established in 1970), Moi University (MU) (established in 1984), Kenyatta University (KU) (established in 1985), Egerton University (EU) (established in 1987), Jomo Kenyatta University of Agriculture and Technology (JKUAT) (established in 1994) and Maseno University (MSU) (established in 2001). The researcher found it necessary to select universities that had been in existence for the longest period because they were more likely to have large accumulations of archives in various formats. They were also presumed to be more advanced in ICT adoption and implementation, hence having more digital
168
records generated, and by extension, digital archives. This selection was done bearing in mind that not all the chosen units had fully functional archival repositories dedicated to the management of non-current records. In multiple case studies, Berg- Schlosser and Meur (2009:114) advised that even though cases should be chosen based on the outcome, cases with both positive and negative outcomes should be selected to enable comparison.
4.4.2 Survey design
Neuman (2014:316) reiterates that the survey design is the most widely used data gathering technique in social science. According to Tella (2015:589), surveys are used in LIS research to gather self-reported data from study respondents. Administering the survey encompasses gathering information ordinarily from fairly big groups of respondents, using questionnaires or structured interview techniques. It sometimes focuses on collecting views of the survey respondents or gathering factual information about the individuals (Tella 2015:589). In the literature, the survey design is lauded for the following reasons:
i. It allows researchers the leeway to collect large amounts of data in relatively short periods of time;
ii. It is a less expensive method of conducting research;
iii. It is easier and faster to administer; and
iv. It can be used to collect information on a wide range of issues (Tella 2015:589).
Nevertheless, Tella (2015:589) pointed out the following shortcomings of surveys:
i. Designing of surveys and their administration can weaken other well- constructed studies;
ii. The respondents’ views and feelings may not be accurately reflected by the alternatives/choices provided as answers in a survey; and
iii. The findings of a survey may be contradicted by the response rate.
There are two main types of survey research design namely the cross-sectional survey and longitudinal survey designs. Cross-sectional design involves collecting quantifiable data relating to two or more variables on more than one case at a single point in time, then analyzing the data to pinpoint patterns of association (Bryman
169
2012:58). In longitudinal design however, samples are surveyed and (re)surveyed again on at least one additional occasion to assess or describe development or change over time (Bryman 2012:63; Cohen, Manion and Morrison 2007:213).
For the present study to be effectively undertaken in the six study sites, the researcher used the survey design. Cross-sectional survey design was applied within the multiple-case (embedded) research design to enable the researcher collect data on a wide range of recordkeeping issues within a shorter period of time, in order to measure and describe the existing digital archiving practices. The survey was carried out by determining the study population, designing, pre-testing and administering questionnaires, carrying out documentary review, conducting interviews and analyzing the collected data. The records and archives management field is not short of success stories on the application of cross-sectional survey designs in MMR such as Laughton (2011), Kalusopa (2011), Luyombya (2010) and Garaba (2010), to mention but a few.