Chapter 3: Research methodology and data collection methods
3.10 Theory testing and building
The theoretical model of systemic institutional design was developed by Healey (2006a, pp.284–310) to manage coexistence within a fragmented society in a collaborative manner. However, this theoretical model was fundamentally developed in relation to the planning context in the West, particularly in the UK (Healey, 2006a, pp.285–286). Arguably, the Jordanian and the British contexts differ significantly, particularly in terms of the social political, economic and cultural context (Meaton and Alnsour, 2012; Abu Dayyeh, 2004; Fischbach, 1994, pp.80–81). Therefore, this research aims to achieve its theoretical contribution firstly by investigating whether this model is applicable and fits the Jordanian context (see the first research objective); and secondly, and more importantly, by examining whether it can be deployed to analyse youth participation in spatial planning in Jordan.
Different social scientists define theory51 from various angles and perspectives. One
school of thought defines theory as the relationships between independent and dependent variables (Colquitt and Zapata-Phelan, 2007). On this trajectory, Campbell defined theory as “a collection of assertions, both verbal and symbolic, that identifies what variables are important and for what reasons, specifies how they are interrelated and why, and identifies the conditions under which they should be related or not related” (1990, p.65, cited in Colquitt and Zapata-Phelan, 2007, p.1281). Other scholars refer to theory as a composition of narratives and accounts. In this vein, DiMaggio defined theory as “an account of a social process, with emphasis on empirical tests of the plausibility of the narrative as well as careful attention to the scope conditions of the account” (1995, p.391, cited in Colquitt and Zapata-Phelan, 2007, p.1281).
However, and before talking about the concepts of theory testing and theory building, it is meaningful to discuss and understand the main elements of theory. Figure 3-3, below, shows the four main blocks of theory. According to Whetten (1989), theory is
composed of four main integrated elements: constructs (concepts); propositions; logic; and boundaries. Constructs are related to the abstract concepts (i.e., culture) by which a chosen phenomenon is explained. Propositions are the logical explanations of how these concepts (dependent variable or independent variable) are linked and related to each other (e.g., cause and effect). Logic is about ‘why’ these concepts are related, while boundaries address issues of ‘who’, ‘when’ and ‘where’ in these concepts and propositions.
Theory testing, as the name suggests, refers to the extent to which evidence from empirical studies confirms (or does not support) an existing theoretical structure. In theory testing, the research uses evidence through specific sorts of data to investigate whether this theory (or model) provides a good explanation for the targeted phenomenon. But before testing theory, it is of great importance to become familiar with all aspects and limitations of the theory to be tested (Bhattacherjee, 2012, p.26; Colquitt and Zapata-Phelan, 2007). Therefore, at the beginning of this study, a critical literature review was conducted, and it is presented in sections 2.3 and 2.4, chapter two. The critical review included deep discussion of the theoretical model (systemic institutional analysis) (see subsection 2.4.1, chapter two) and the theoretical
Figure 3-3: The concept of logical consistency in theory testing, source: Bhattacherjee, 2012, p.28, drawn by the author
underpinnings of the institutionalist and collaborative planning approach (see subsections 2.3.1, 2.3.2 and 2.3.3, chapter two). Discussing the dimensions of the institutionalist and collaborative planning approach led to greater understanding of the theoretical model and thus to deploying it to frame the empirical data collection. In the vein of theory testing, Bhattacherjee (2012, p.28) provided different means to differentiate a good theory from a weak one. Bhattacherjee’s (2012) attributes were devised to test the adopted theoretical model. The four suggested methods are as follows:
logical consistency falsifiability parsimony
explanatory power.
Logical consistency determines whether the main components of theory (constructs, propositions, logic and boundaries) are all intermeshed and integrated with each other (see figure 3-3, above). Falsifiability means a theory is not a good theory unless it can be tested empirically. Parsimony refers to the simplicity of a theory. In this regard, simplicity means the best theories are the simplest ones. However, the method of theory testing most commonly used in the literature is that of measuring a theory’s explanatory power. Under this category, a good theory explains (or predicts) reality or a target phenomenon better than rival theories.
This research relies on the fourth category to test the theoretical model. This study measures the extent to which systemic institutional design explains and investigates the phenomenon of youth participation in spatial planning. In this qualitative research, the strength of this theoretical model was tested through pattern matching, by observing the extent to which the concepts and patterns embedded in the theoretical model match the concepts and patterns emerging from the empirical data (see figure 3-2, above). Those patterns and concepts that do not match or fit the model but were repeatedly pointed out and labelled ‘critical’ in the raw research data and they generated a new theme named cultural capital. The process of expanding/modifying a theoretical structure has been tagged in the literature as ‘theory building’ and is explained as follows.
Eisenhardt and Graebner (2007) and Eisenhardt (1989) stated that building theory could involve the use of empirical evidence from one or more case studies. According to Steinfield and Fulk (1990, pp.16–18), four main techniques were presented in the literature to ‘build’ a theory:
ground theory building
bottom-up conceptual analysis
applying existing theories in new contexts extending or modifying existing theories.
Ground theory involves an “iterative process of collecting and [analysing] data in order to build a theory about how actors interpret their daily realities)” (Colquitt and Zapata-Phelan, 2007, p.1282). However, one of the main criticisms of Ground theory is that the resulting theory is subjective and non-confirmable (Bhattacherjee, 2012, p.28; Eisenhardt and Graebner, 2007). The principles of Ground theory were used in this research to generate and validate concepts, patterns and themes from primary data through open coding, axial coding and selective coding. As previously discussed, it was never the intention of this study to generate a new theory. However, this research was intended to extend/modify the theoretical model (systemic institutional design) to make it more applicable to the Jordanian context.
The second technique, bottom-up conceptual analysis, is about building theories through “[identifying] different sets of predictors relevant to the phenomenon of interest using a predefined framework” (Bhattacherjee, 2012, p.29). In this approach, there is a defined model or framework with a simple input–process–output framework. Researchers who adopt this approach try to figure out how the inputs, processes and outputs are connected to each other.
The first two approaches are inductive in nature and rely heavily on the researcher’s abilities to interpret the data being studied. According to Bhattacherjee (2012, p.29), prior knowledge of the theoretical model may lead to biased results in these two inductive approaches. As this research started with extensive reading and a critical review of the theory of the institutionalist and collaborative planning approach, it is believed that the first two inductive approaches are incapable of achieving theory building and thus dismissed as techniques for theory building.
The third approach refers to applying existing theories in new contexts. Applying existing theories in new contexts has a high level of similarity to the approach of extending or modifying an existing theory (the fourth approach), as both approaches are deductive in nature. The researcher chose a particular theoretical model and placed it in a completely new context (the Jordanian context). In this model, the researcher was trying to explore ‘constructional’ similarities between the two contexts. However, such approaches are more applicable to contexts with high structural similarities than those with radical and structural differences, as with British and Jordanian planning practices (Meaton and Alnsour, 2012; Abu Ghazalah, 2010, 2008a; Abu Dayyeh, 2004; Fischbach, 1994, pp.80–81).
The fourth approach is about extending or modifying an existing theory. This study followed this approach to explain a specific phenomenon (i.e., youth participation in spatial planning) in its context (the Jordanian context). In this approach, the researcher refers to certain pre-identified concepts, propositions or boundary conditions of the ‘old’ theoretical model to retain/modify/extend the model to make it fit the new context better. In this way, a better explanation of the target phenomenon would be achieved. To put it simply, the constraints and challenges that define the theoretical model (systemic institutional design) adopted to achieve collaborative planning are primarily developed according to the Global North context, and particularly according to the UK context (Farzaneh, 2011, p.20). However, this research is trying to adapt this very theoretical model to achieve the theoretical contribution of the research. According to the institutionalist paradigm, the barriers that hinder youth participation in the UK are different from the structural or embedded barriers in the Jordanian context. Yet there are some structural similarities between both contexts in terms of planning systems. At the inception of the Jordanian state, the governance body of Amman relied to a large extent on the British Town and Country Planning Act of 1947 to develop its embryonic planning system (Meaton and Alnsour, 2012; Abu Dayyeh, 2004; Fischbach, 1994, pp.80–81). Both contexts encompass highly complex bureaucratic planning institutions (Harris, 1958, p.87). As discussed in section 2.2, chapter two, scientific rationalism in policy analysis assumes that the state is capable of making public policy and ‘controlling’ the locations of development single-handedly (Healey, 2006a, p.99). Therefore, the ‘old’ model is to some extent applicable to the Jordanian context. However, planning practices in the two contexts to a large extent differ from each other
in social, cultural and political terms (Al Rabady et al., 2014). Also, the subjects of plans differ significantly. British planning practices are more directed towards urban regeneration and the promotion of sustainable urban development (Vigar et al., 2000, p.10), in contrast, the Jordanian plans are more about regulating urban expansion and zoning in principle, encompassing building density and the zoning of new urban land (Tewfik and Amr, 2014; Meaton and Alnsour, 2012; Haddaden, 2009, p.25; Potter et al., 2009; Bagaeen, 2006a).
Colquitt and Zapata-Phelan (2007) provided a taxonomy to level the theoretical contribution of an empirical study (see figure 3-4, below). On the vertical axis of the chart, Y1 and Y2 represent a low level of theory building. Replications of previous findings do not contribute to existing theory, nor do they introduce new relationships or constructs. They build on previous findings to confirm new relationships between concepts already described in past research. Y3 represents a moderate level of theoretical contribution because it only clarifies existing theory. At this level, empirical studies may supplement a new substantive mediator or moderator of an existing relationship or process. Levels Y4 and Y5 represent a high level of theory contribution. Empirical studies rated as Y4 or Y5 have investigated and introduced previously unexplored relationships or processes of the theoretical structure. Ideas presented in an empirical study that are rated Y5 represent a radical departure from the extant literature. This study has brought new conceptual thinking into the theoretical model and thus it belongs to level Y4. The introduction of the theme of sociocultural forces into the model is a departure point for generating a number of new research directions that can shape future thinking on youth empowerment policies in Jordan.
The horizontal dimension of the chart measures theory testing. X1 and X2 both represent low levels of theory testing. Mostly inductive studies rated X1 employ the hypothetical-deductive model and are low in theory testing. For X2, empirical studies rely on the extant literature to ground a priori hypotheses. According to Sutton and Staw, referencing past findings with no logical explication is no more than “a smoke screen to hide the absence of theory” (1995, p.373, cited in Colquitt and Zapata-Phelan, 2007, p.1285). However, if an empirical study succeeds in convincing reviewers of the explication of all the causal logic on which predictions are based this may result in a moderate level of theory testing (X3). Nevertheless, predictions of empirical studies
rated X3 have not been developed based on existing theory, nor do they deeply examine the target phenomenon. X4 and X5 represent a high level of theory testing. Research questions in this study are grounded in an existing theoretical model to examine the phenomenon of youth participation. This implies that it falls at the level of X4 in terms of theory testing. Accordingly, this study comes closer to theory testing. However, a theoretical model is only a symbolic representation of the original theory. All studies that ground their prediction in existing theory are at the highest level in theory testing. According to Colquitt and Zapata-Phelan (2007), empirical studies offer strong theoretical contributions if they offer strong theory building or/and strong theory testing. Studies at this level are termed by Colquitt and Zapata-Phelan, in figure 3-4, below, as the expanders. This has served as the foundation for new theoretical constructs in the adopted theoretical model of systemic institutional design (level Y4, according to figure 3-4, below). Also, this design of research objectives and related research questions was grounded in the theoretical model of systemic institutional design (X4).
`
Y axis: Building a new theory
Y5: Introduces a new construct (or significantly
reconceptualises an existing one)
Y4: Examines a previously unexplored relationship or process Y3: Introduces a new mediator or moderator of an existing
relationship or process
Y2: Examines the effects that have been the subject of prior
theorising
Y1: Attempts to replicate previously demonstrated effects X axis: Testing existing theory
X5: Grounds predictions in existing theory X4: Grounds predictions in existing models
X3: Grounds predictions in existing conceptual arguments X2: Grounds predictions in references to past findings
X1: Is inductive or grounds predictions in logical speculation?
Figure 3-4: Taxonomy of theory building and theory testing, source: Colquitt and Zapata-Phelan, 2007, p.1283
Y1
Y4
Y3
Y2
Y5
X1 X2
X3
X4
X5
Builders Reporters Qualifiers Expanders Testers Above the arc: High theoretical contributionUnder the arc: Low theoretical contribution