3.1 Overview

Marcos Antonio de Lima Filho, PhD.

This study proposes an evolutionary theory for the evolution of designs. This contribution is the product of a grounded theory methodology (Sections 3.3 and 3.4). The following components were incorporated into this research methodology: the study employed a comparative analysis (I) of two diverse industries through a census of commercial aircraft and a census of smartphone specifications (II). The census data were used to build a historical time series (III), covering a period of 88 and 20 years, respectively, of technological evolution. The research mixed quantitative and qualitative data (IV), to inquire about an issue that emerged from data (V). Some of these elements have been adopted by similar research (VI) in various fields: censuses have been employed in seminal studies of industrial evolution, disruptive innovation, as well as ecological studies.

The next paragraphs provide an overview of the research design, briefly summarising the topics enumerated above. These themes are discussed in more detail throughout the chapter.

I. Comparative Analysis

This grounded theory adopted a “most different” type of comparative analysis. To do this, the researcher purposefully selects cases that differ as much as possible, since the research objective is to find similar processes in diverse settings (George & Bennett, 2005). When comparing different cases, such as commercial aviation and consumer electronics, “the researcher is forced to distil out of that diversity a set of common elements with great explanatory power” (Collier, 1991). This type of comparison can greatly aid the researcher in transcending substantive descriptions of time and place as he or she tries to achieve a general, formal theory (Glaser & Strauss, 1967/2006, p. 55). As a result, the comparison of “most different” cases probably has stronger claims of representativeness and external validation than other small sample designs (Seawright & Gerring, 2008).

II. Census

Surveys are often contrasted with censuses, and both employ many of the same methods. However, whereas a census is intended to gather information about all members of a population of interest, a survey gathers information from only some of the population members, that is, from a sample of the population (Lavrakas, 2008). While the most commonly recognised type of census focuses on demographic data, censuses can also be used to gather information on economic activities and trends (Cantwell, 2008).

III. Historical Time Series

Archival data sources can enable a complete census of population members over time and the identification of performance changes in key technological parameters (Anderson & Tushman, 1990). With this data, the researcher can then visualise a historical time series of how an industry evolved. Time series analysis can unveil the progression of a technology over time, providing valuable insights into a range of questions, such as which features are being embraced or rejected, what is experiencing growth or decline, how competing technologies are performing, the level of disruption caused by the technology, which technologies are being displaced, and whether it represents a dominant design.

IV. Quantitative and Qualitative Data

Glaser (1992, 1998, 2003) asserts that grounded theory is neither qualitative (which leans towards constructivism) nor quantitative (which inclines towards objectivism); rather, it is a distinct, general, inductive, concept/theory-generating method that can draw from both but is not confined to either, unless one insists on a dualistic understanding (Martin & Gynnild, 2011). Therefore, when developing a grounded theory, it is essential to consider not only qualitative data but also quantitative data, as these can contribute to refine and generate new theories (Walsh, 2015).

V. Preconception vs Emergence

In grounded theory research, the research design is not supposed to be planned but rather to emerge through constant comparative analysis and theoretical sampling (Johnson & Walsh, 2019). Contrary to the traditional quantitative approach, which implies the articulation of hypotheses in advance, or the traditional qualitative approach, which implies a theoretical framework to be applied, Grounded Theory does not require as a starting point the identification of a gap in the literature from which precise research questions are framed to guide the study (Holton & Walsh, 2017).

It is important to state such distinction because “the precepts of traditional qualitative research are often misaligned with grounded theory’s precepts” (Holton & Walsh, 2017). Grounded theory is a structured and systematic methodology, on par with any qualitative method, figuring among the most widely used research methods. The main difference relies on the preconception vs emergence of its components: research questions, hypothesis, theoretical frameworks, and the use of extant literature. For most methods, these are usually set a priori. However, for Classic Grounded Theory, these components emerge throughout the research process.

VI. Similar Research

Similar innovation studies have incorporated historical time series analysis, comparative designs, and censuses. Abernathy and colleagues pioneered the research of industrial evolution, dominant designs, and disruptive innovations. These concepts emerged from a historical analysis of the US automobile industry (Abernathy et al., 1983; Abernathy & Clark, 1985). Christensen (1992a; 1992b; 1993) also conducted a historical time series analysis of the hard disc-drive industry. To study technological evolution and radical innovation, Sood and Tellis (2005) also collected data using the historical method.

The comparative analysis of diverse industries has a long history in this research tradition. Tushman and Anderson (1986) compared three different sectors: “Using data from the minicomputer, cement, and airline industries from their births through 1980, we demonstrate that technology evolves through periods of incremental change punctuated by technological breakthroughs”. Nair and Ahlstrom (2003) also employed a “most different” type of comparative analysis, sampling two disparate settings: steel manufacturing and chronic kidney disease treatment. Suárez and Utterback (1995) compared six industries: the automobile, typewriter, transistor, electronic calculator, television, and picture tube industries. Their time series extends back to 1874.

Censuses are seldom used in innovation studies due to their larger scope and the challenges of gathering data about all elements of a population. This approach poses significant challenges for researchers in terms of time, resources, and data accuracy. Given these obstacles, I have found only a limited number of innovation studies that have made use of censuses (Table 3.1.1). However, when the target population is small, a census study may become feasible and even necessary: when the population is small and variable, any sample we draw may not be representative of the population (Cooper & Schindler, 2014).