100%(4)4 out of 4 people found this document helpful
This preview shows page 49 - 52 out of 230 pages.
2012). The automated process helps mitigate the common problem of subjective biases with a manual coding process. This technique not only counts the frequency of occurrence of a keyword or a string of keywords, but also assesses the interconnectedness of keywords in the document based on network analysis (McPhee, Corman, & Dooley, 2002; Hofer et al., 2012). Keywords with many connections to other words may be described as “central”. In other words, the more words that connect to a particular keyword, the greater the “betweenness centrality (BC)” of that keyword. The focus of this CRA technique is to identify those keywords that have high BC scores as measured by the influence level of keywords in a text. Mathematically, the influence (I) of a keyword in a text T is represented using a social network metric as follows (Corman, Kuhn, McPhee, & Dooley, 2002): g1835g3036g3021g3404 ∑ g1859g3037g3038g4666g3036g4667/g1859g3037g3038g3037g2996g3038g4670g4666g1840g33981g4667g4666g1840g33982g46672g4671where
49 g1835g3036g3021= influence of a word iin text Tg1859g3037g3038= number of shortest paths connecting jthand kthwords g1859g3037g3038g4666g3036g4667= number of those paths containing word i N = numbers of words in the network Crawdad also calculates the correlation value between two words. A positive correlation between a pair of words suggests that the given pair tends to co-occur in close proximity in the text (Hofer et al. 2012). This is defined mathematically by Corman et al. (2002) as follows g1842g3036g3037g3021= g1835g3036g3021. g1835g3037g3021. g1832g3036g3037g3021where g1842g3036g3037g3021is the correlation value between words g1875g3036g3021and g1875g3037g3021g1835g3036g3021is the influence of word g1875g3036g3021g1835g3037g3021is the influence of word g1875g3037g3021g1832g3036g3037g3021is the number of times that g1875g3036g3021and g1875g3037g3021co-occur (their corresponding nodes are connected directly by an edge) in text TThe objective of structured content analysis in this research was to pull out relevant excerpts from company reports that reflect various themes and strategies related to market-oriented supply chain sustainability. In order to achieve this objective, only those excerpts had to be extracted which reflected keywords related to market-orientation and supply chain management, and also contained most influential words in the reports. The process adopted to achieve this objective is described as follows.
50 Keywords related to market-orientation as conceptualized by Slater and Narver (1995) and Matsuno and Mentzer (2000) were shortlisted. These keywords are: market, employee, customer, staff, stakeholder, passenger, people, society, shareholder, investor, government, supplier, and competitor. The keywords representing supply chain management were shortlisted as per Rossetti and Dooley (2010) who provide a list of keywords associated with supply chain management. These keywords are: operation, network, supply, chain, source, management, transport, schedule, quality, procurement, purchasing. The keywords representing sustainability and strategy were not included in the matrix because the reports being analyzed in this research are by definition “sustainability reports” and we are arriving at strategies from these reports.