Chapter 3 Research design
3.6 Case research design
Many scholars stress that after choosing case study research as the research method and before data collection and analysis, the researcher needs to carefully consider case research design from three aspects: unit of analysis, single- or multiple-case research, and case selection (Benbasat, Goldstein, & Mead, 2002; Yin, 2009). Failing to do so, say those scholars, will undermine the quality of data collection and analysis, subsequent inferences and conclusions, and the validity and reliability of the research.
3.6.1 Unit of analysis
The first aspect to consider in designing case study research pertains to unit of analysis. It denotes the research focus and therefore can be individuals, small groups, or organisations; in some circumstances, it can also be relationships or decisions (Benbasat et al., 2002; Yin, 2009). Guest, Namey, and Mitchell (2013, p.
26) add that unit of analysis ‘is the level of abstraction at which you look for variability’. The research questions are central in defining the unit of analysis (Benbasat et al., 2002; Yin, 2009).
The unit of analysis in this study is defined as political parties’ campaign teams, in other words, small groups, because this entire study is concerned with political parties’ e-campaigning conducted by their campaign teams in the 2008 New Zealand general election.
3.6.2 Single-case or multiple-case research
Another important aspect of case research design is the number of cases to be conducted or studied; in other words, the researcher needs to decide if the empirical research should be based on only one case or multiple cases (Benbasat, et al., 2002; Yin, 2009).
A single case design is justified when the case ‘represents the critical case in testing a well-formulated theory’ that ‘has specified a clear set of propositions as well as the circumstances within which the propositions are believed to be true’
(Yin, 2009, p. 47); represents an extreme case or a unique case that ‘commonly occurs in clinical psychology, where a specific injury or disorder may be so rare that any single case is worth documenting’ (Yin, 2009, p. 47); is ‘the representative or typical case’ and ‘the objective is to capture the circumstances and conditions of an everyday or commonplace situation’ (Yin, 2009, p. 48); is ‘the revelatory case’
that ‘exists when an investigator has an opportunity to observe and analyse a phenomenon previously inaccessible to social science inquiry’ (Yin, 2009, p. 48);
or is ‘the longitudinal case: studying the same single case at two or more different points in time’ (Yin, 2009, p. 49).
A multiple-case design is advised when the rationale for single-case research cannot be satisfied; that is, the case is not critical, extreme, representative, revelatory, or longitudinal (Benbasat et al., 2002; Yin 2009). Yin remarks that a case is analogous to an experiment in quantitative research. Thus, ‘the evidence from multiple cases is often considered more compelling, and the overall study is therefore regarded as being more robust’ (Herriott & Firestone, as cited in Yin, 2009, p. 53). Benbasat et al. (2002, p. 84) agree and add that a multiple-case design allows for ‘cross-case analysis and the extension of theory’. Yin (2009, p. 54) further indicates that a multiple-case design enables the researcher to replicate the same analytic logic – as opposed to sampling logic in quantitative surveys – as
in the first case, in order to predict similar results (a literal replication) or predict
‘contrasting results but for anticipatable reasons (a theoretical replication)’. Yin (2009, p. 53), however, warns that ‘the conduct of a multiple-case study can require extensive resources and time beyond the means of a single student or independent research investigator’.
This study opts for a multiple-case design because the rationale for a single-case design cannot be fulfilled. Further, multiple case studies enable more compelling empirical evidence, a cross-case analysis for further insights, and theoretical extension and replication.
3.6.3 Case selection
Case selection is central in any case research design (Benbasat et al., 2002;
Seawright & Gerring, 2008; Yin, 2009). However, specific advice concerning this aspect is sparse in research methods literature (Seawright & Gerring, 2008).
Benbasat et al. (2002, p. 85) stress that case selection ‘should be carefully thought out rather than being opportunistic’. Yin suggests that in multiple-case designs, researchers should consider two or three cases for literal replications and four to six cases for theoretical replications. Seawright and Gerring point out that in multiple-case designs researchers base their case selections on various considerations. For instance, some researchers choose cases primarily for pragmatic reasons, ‘such as time, money, expertise, and access’ (Seawright &
Gerring, 2008, p. 295); some choose cases based on random sampling; and some choose cases because of diversity.
In this study, case selection is based on three main criteria. First, the cases should reflect diversity; in other words, they should not focus on only the major parties’
or minor parties’ e-campaigning utilisation. Diversity is also a main characteristic of New Zealand’s political landscape, especially after MMP replacing FPP as the electoral system for general elections in 1996 (Miller, 2010; Mulgan, 2004).
Second, the cases should be concerned with only the parliamentary parties because they had a more realistic chance of being elected and thus had more meaningful election campaigns in general. It is worth noting that most existing e-campaigning studies based on party-led elections solely or chiefly focus on
parliamentary parties (e.g., Gibson & McAllister, 2008; Schweitzer, 2008a; Small et al., 2008; Voerman & Boogers, 2008). And third, the cases should each feature e-campaigning utilisation or it would render this study meaningless.
Eight parliamentary parties participated in the 2008 New Zealand general election:
the New Zealand National Party (referred to as ‘National’ henceforth), the New Zealand Labour Party (referred to as ‘Labour’ henceforth), the Green Party of Aotearoa (referred to as ‘the Greens’ henceforth), the New Zealand First Party, ACT New Zealand (referred to as ‘ACT’ henceforth), the Maori Party, Jim Anderton’s Progressive Party (referred to as ‘the Progressives’ henceforth), and United Future New Zealand (referred to as ‘United Future’ henceforth). Those parties all conducted e-campaigning in the election. Thus, there were eight cases satisfying the three criteria for case selection.
The researcher contacted all case organisations, inviting their campaign teams to participate in this study. Seven accepted, including the major parties; one minor party, however, indicated that the person solely responsible for its e-campaigning passed away at the early phase of the campaign period and the party’s e-campaigning had remained mostly unattended since. Accordingly, this study investigated the e-campaigning utilisation of six parliamentary parties, namely, National, Labour, the Greens, ACT, the Progressives, and United Future. Although the e-campaigning of two minor parties are not included in this study, the included cases still highly reflect diversity.