Home Business & Finance
Table of Contents:
Thematic data analysis is a flexible research tool, which provides a rich, complex, and detailed account of the data (Vaismoradi, Turunen, & Bondas, 2013). It is well-understood in the qualitative research realm as it is one of the most common qualitative analysis methods. Braun and Clarke (2006) describe thematic analysis as an approach to recognize, analyze, and report patterns identified within the data. These identified patterns/themes provide insight of the phenomenon being studied. Thematic analysis pulls out the patterns to generate the theory (Braun & Clarke,
2006). This type of analysis is intended to provide a transparent view of the data and what the data represent in terms of the real world of IT hiring managers trying to fill data scientist/big data specialist roles.
Structural analysis provides a sequential methodology which is particularly applicable to interviews (Herz, Peters, & Truschkat, 2014). Similarities exist between structural analysis and thematic analysis in that they both employ condensing data, coding, and analyzing them. Structural analysis aims to analyze both the inside perspective and the external attributes those perspectives reveal. This approach is chosen to reveal structured/sequential insight into the data, which can be applied against the patterns identified in the thematic analysis approach.
A key point of structural analysis is theoretical sensitivity. Theoretical sensitivity involves the researcher’s capability to distinguish pertinent from nonpertinent data to develop a theory that is pragmatic and well-integrated. This is a disciplined approach to help create theories from data. Where thematic analysis can identify several patterns in the data, structured analysis aims to find sequences, or structured processes, in the data (Herz et al„ 2014).
Data saturation assures that the themes throughout the study are well-supported, and no new information exists over each topic that would change the researcher’s understanding of the concept. While the survey questions were designed to collect information about the big data resource achievement steps, the interview questions for IT Managers were designed to gather additional details to understand similarities and differences between the same or similar survey responses and confirm if saturation was, indeed, reached.
Big Data Initiatives
A key area to understand about the big data landscape is what are businesses doing in the big data space. As explained earlier, predictive analytics can provide a new level of benefit to individuals and organizations. While still in its early stages, organizations are still building their big data environments and working on achieving new insights from the new types of data they are collecting. With 63 big data initiatives captured in the survey from 20 research subjects, Table 10.2 lists these initiatives in order of number of responses.
Big Data Initiatives
Years of Big Data Initiatives
To learn how long the subjects’ organizations have been working on big data initiatives, it was asked of them to share the year they began, in whatever capacity, their path down the road of big data analytics. Table 10.3 shows the time frame of the beginning of the big data initiatives from the twenty research subjects by listing the years that the big data efforts, most current first, as well as the number of responses for those years.
Size of Big Data Teams
To understand how large the big data specialist teams are in the organizations being represented in the research, Table 10.4 reflects the size of the big data teams at the subjects’ organizations.
Big Data Resources Needed
The term “Data Scientist” is ambiguous in that it may encompass someone who deals with big data in a limited or larger number of ways. A data scientist can be someone who knows and can perform one or more of the following: collect and manage data, build the infrastructure to hold the data, understand and apply math and statistics, develop algorithm and code, analyze results, and communicate with business leaders to gather priorities and report findings. Some organizations have different titles for each of those individual roles, such as developers, data miners, engineers, architects, evangelists, etc.
Table 10.5 illustrates the resources that the organizations in the research subjects represent are still need to be found. “Data Scientist” was specifically mentioned, as
Time-frame Respondents Started Their Big Data Initiative
Size of Respondents' Big Data Teams
were roles that the subjects pointed out that were particularly needed. Integrators, data architects, and engineers were categorized under the heading, “technologists.” Needed roles that were specified in the data w'ith the terms “Chief,” “Lead,” or “Head” were counted under the category titled, “Leadership.”
The distinction between “Technologists” and “Developers” may not be clear. While the technologist’s roles are to make the hardware and software function, developers create the code that addresses the data sets to find value. Technologists are often considered part of the infrastructure or operations group, and developers write code that will run on that infrastructure.
Where Are Organizations Finding Big Data Resources?
The key question of the research is “Where are organizations finding their big data resources?” The 20 research subjects provided a total of 57 answers of the places they turned to find big data resources. Table 10.6 breaks dowrn these places based on the frequency, w'hich they were answered:
Big Data Resources Still Needed
Where Respondents Found Big Data Resources
The 57 answers summarized into twelve categories for Table 10.6 came from four separate questions, asking the subjects to rank their answers in terms of most effective and list the most effective first, followed by the second most effective, the third, then the fourth. Nine of the “most effective” responses from each subject were “Internal” and five were “Referral”; the two most successful areas where the subjects’ organizations found their big data resources.
Challenges Finding Big Data Resources
The challenges in finding big data resources listed by the 20 research subjects fell into five different areas. Do the candidates have the skills or experience? If so, can the organization afford them? Are they looking in the right places? Table 10.7 lists each of those areas along with the number of responses:
Qualities Most Difficult to Find in Candidates
Identifying the qualities that are most difficult for organizations to find in their big data specialist candidates fell into four different categories. A category titled “Personality Qualities” included those soft skills with comments that included the ability to communicate effectively, work well in a team environment, drive, tenacity, and an ability to understand the big picture of what the use of data science is all about in the context of the business. Table 10.8 presents the qualities IT managers are seeking in the order in which they were most identified by the survey subjects:
Challenges Finding Big Data Resources
Qualities Most Difficult to Find in Big Data Candidates
The Ideal Big Data Specialist Candidate
Research subjects were asked to provide a description of their ideal big data specialists candidate. Most of the answers included multiple characteristics. As described above, there were personality qualities several of the respondents mentioned, including ability to teach others, desire to learn new' things, emotionally intelligent, ability to tell a story, can deduce, or has leadership skills. These have been grouped into a category titled “Personality Qualities.” Table 10.9 lists the ideal attributes and the number of responses to each:
Number of Candidates Interviewed
The number of candidates interviewed by the businesses represented in the research number into the thousands. Not every company though has a large number of big data specialists/data scientists. Table 10.10 lists the number of big data candidates interviewed listed by range of interviewees followed by the responses from the research subjects:
Easing the Big Data Hiring Process
The final question of the survey sought to learn what could be done to make the big data resource hiring process easier or more successful. Six categories were identified out of 40 responses. Table 10.11 lists those categories, sorted by the number of responses that mention them.
“Better Sources of Candidates” covered such areas as improved networking, more qualified resumes, prevalidated candidates, improved use of internal candidates, etc. This w'as the largest area identified by the research subjects to improve the hiring of big data candidates.
The Ideal Big Data Candidate's Qualifications
Number of Candidates Interviewed
Areas to Improve the Big Data Hiring Process
The term “Adaptability” covered an assortment of areas ranging from allowing employees to work remotely, greater flexibility on behalf of the hiring company with compensation, or less limitations due to non-US Citizen candidates. Seven of the subjects mentioned that regardless if it was the candidates themselves or the organization that was looking to hire someone, being more adaptable in these areas would make the hiring process easier.