Qualitative analysis technique
Qualitative techniques are best used to analyze business data because they discern, examine, compare, contrast, and interpret meaningful patterns and themes. Data obtained is analyzed and synthesized through various angles according to the research questions given. There are many techniques used to analyze data quantitatively, they include ethnography, discourse analysis, textual analysis, typology, taxonomy, constant comparison/ grounded theory, analytic induction, logical analysis/matrix analysis, quasi statistics, event analysis/microanalysis, metaphorical analysis, domain analysis, hermeneutical analysis and narrative analysis (Ratcliff, n.d). They all correspond to various objectives, philosophical orientations, disciplinary traditions and types of data.
Qualitative analysis is considered because it is systematic and follows discipline. It considers various assumptions and processes and consumes less time. Qualitative analysis is divided in small stages which involve development of instrument, collection of data, processing of data and analyzing data which makes it an interactive process. Qualitative analysis examines collected data and determines the process it uses to answer evaluation questions provided. Qualitative data occurs in reducible and embedded forms compared to quantitative data. the process of data analysis includes data reduction, data display, drawing conclusion and verification.
Analytic induction method uses raw data readings that are detailed to derive themes, concepts and models. It is among methods used to analyze business data as a qualitatively. It discovers concepts and relationships between them through use of collected data. researchers using analytical induction gather data from many situations, examines it to find the data that is suggested by the collection (Degroote school of business, 2002). The method defines and uses terms before conducting a research whereby the definitions given are used as hypothesis ready to be tested. It is inductive rather than being deductive. It involves reasoning, concept modification and offers relationships to concepts in research process. It has a goal of giving accurate information to create a reality of the situation (Katz, 2015). The research makes universal statements that are subject to modification if there are exceptions discovered. The goal of the obtained knowledge is causation, which may include many exceptions. It analyzes perceptions of end-users together with their attitudes towards their participation in virtual communities.
A careful examination of similarities is done in social phenomenon to develop ideas and concepts. The method induces laws from analysis of instances that are isolated and modifies them through research, which allows accurate study of the phenomenon studied. Analytic induction has its bases on ontological and epistemological realism that depends on Cartesian dualism and assumes the truth and its equation with faithful reproduction of the object as part of real knowledge (Thomas, 2006). Analytical induction is most considered because it encourages self-awareness and regard. It promotes motivation, emotions and desire and compulsion to act.
It is inductive and not deductive
It fits well when viewed from ethnographic point.
It introduces observations to develop better hypothesis through permitting revisions
It redefines the studied content to eliminate exclusions.
It does not realize universalism initially
It focuses on causation, provisionally generates and tests theories, restricts all data to be tested using hypothesis. Its grounded theory concentrates on comparison and generation.
According to Robert (2003), analytic induction permits inference because it covers many topics which are discussed separately. These are analytic induction, validity and reliability, quantification, sampling, generalizing and argumentation.
Steps in Analytic induction process
Analytic induction process involves several steps such as
- Definition of a phenomenon in a tentative way. this involves analyzing small cases closely. Hunches and inspirations are pushed further to encourage creativity.
- Development of hypothetical statement about the phenomenon from the analysis.
- Consideration of single instances to confirm the hypothesis. This involves testing of hypotheses with the similar data (Onwuegbuzie, Leech, & Collins, 2012).
- Failure to confirm the hypothesis the phenomenon is defined again to include examined instances. The process is repeated until a single hypothesis passes the preliminary test. All negative cases are analyzed and those that do not fit are discarded.
- More cases are examined to determine the hypothesis that is confirmed repeatedly which forms a certain degree of certainty regarding the hypothesis.
- In case there are negative cases the hypothesis are formulated again until all exceptions are eliminated
- Comparison of hypothesis with alternative possibilities obtained from other circumstances. The process continues until all cases have been analyzed and a description that describes all data is attained (Robert, 2003). The process achieves a conceptual framework arranged from more important concepts to less important ones. Otherwise referred as an interpretation. The final step involves generalization of interpretation with comparative data from other researches and studies done and personal knowledge of the world. The end result is an analysis that is reliable and that which covers most parts of data and most variations of the phenomenon studied.
Analytic induction begins with data collection. The process involves selecting certain locations to collect data, focusing on the elements, simplifying them, abstracting and transforming the collected data into notes and transcriptions. The process is also referred as data reduction because it condenses collected data for easier management of the issues to be addressed. Data reduction gives options about data to be emphasized on, minimized, and set aside for the project.
The second step is to develop analysis of the gathered data. This process includes display of gathered data in form of charts, diagrams, and matrixes which offers new ways of arranging and thinking about textually embedded data (Ratcliff, n.d). Analysts extrapolate totally from the reduced data and discern systematic patterns and their interrelationships. At this stage, there emerge more categories and themes from the data that is beyond that which was initially discovered during data reduction. In this step analysts identify reasons behind working of some projects and not others. Example analysts may display narrative data by developing flow charts to map critical paths, supporting evidences and decision points (Miller, 2008). The second process involves repetition for the remaining areas. Analysts can use subsequent site data to change the flow chart that was original, prepare independent flow chart in each site and prepare single flow chart for all sites and events.
The final step involves presentation of research findings. The main objective of Analytic induction is causal explanation of the phenomenon for explanation and its explanatory factors in that a perfect relationship is maintained. Initial cases are inspected to locate common factors and provide explanations. At this stage original hypothesis are contradicted and their explanations reworked. Definitions of explanandum are redefined to eliminate cases that are troublesome or to get them consistent with explanas (Smelser and Baltes, 2001). The explanas are revised to make all target phenomenon cases show conditions that are well explained. Analysts seek encounters with new data so that they can make any revision that makes valid analysis in many diverse cases.
In Conclusion drawing and verification is the last step in analytical induction. It includes stepping back to get the original meaning if the analyzed data and measure its implication for questions given. At this stage data is verified as many times as possible to cross-check conclusions that are emergent (Gibbs, 2011). Data is tested for its sturdiness, plausibility, conformability among other factors. Data is tested for its validity which examines whether conclusions made are warranted, defensible and credible. They are also examined whether they can withstand the alternate explanations. Investigation continues to a point when analysts cannot pursue negative cases.
Features of categories developed from coding
Outcomes of inductive analysis lead to a development of categories into frameworks that sum up raw data, key themes and processes. There are five features of results that are obtained from coding categories. These include category label, description, texts, data, and links.
Category label includes short phrases and words that refer to that category (Thomas, 2006). The labels bear inherent meanings that reflect main features of that category. Category description includes a description of that category and its meaning. In addition, there are key characteristics, limitations, and scope of that category.
Category text and data is coded to show associations, perspectives, and meanings. Each category also has links with other categories. The links indicate super ordinate, subordinate and parallel categories. The category is also embedded in a model, framework and theory. Frameworks may include open network, which does not have a sequence, a temporal sequence that includes movement over time and causal network where one category leads to changes in another (Thomas, 2006).
Coding schedule, coding manual, coding units, further explanation; provide examples; sample of texts
Coding involves combination of themes, categories, and ideas, which helps in forming similar text passages with their labels that can be retrieved at later stages for comparison and analysis. Codes are based on topics, themes, concepts, ideas, phrases, keywords, and terms.
When coding researchers look for the processes going on, what is being done and said. What is supported by structure, contexts, and the changes made.
Examples of what to be coded include behaviors and actions of people seeking reassurance and bragging. Events such as weddings and first day people begin working. Activities like clubbing, attending night courses and conservation works. Strategies when people get nasty and deserve to be fired from jobs. States such as when workers felt good and hopeless. Participation with other employees, relationships and interactions with other members. Consequences of certain actions like positive attitudes and opportunities.
Degroote School of Business. (2009). Analytic induction & Abductive reasoning modes for theory generation in the qualitative phase of investigation, ph.D.Dissertation,1-5.
Gibbs, G. (2011). Learning qualitative data analysis on the web. Online QDA. Retrieved from http://onlineqda.hud.ac.uk/methodologies.php
Gibbs, G.,& Taylor, C.(2010). How and what to code. Learning qualitative data analysis on the web. Retrieved from http://onlineqda.hud.ac.uk/Intro_QDA/how_what_to_code.php
Hicks, A. (2006). Qualitative comparative analysis and analytical induction. The case of the emergence of the social security state. Sociological methods research, 1(23), 86-113.
Katz, J. (2001). Analytic induction. International encyclopedia of the social & behavioral sciences, 480-484.
Katz, J. (2015). Superiority and banality of the qualitative method; the analytic induction of jack Katz, 2(1),147-154.
Miller, S. (2008). Quality and quantity; another view of analytic induction as a research technique. Quality and quantity,4 (16),281-295.
Onwuegbuzie, A., Leech, N., Collins, K. (2012). Qualitative analysis technique for the review of the literature, 56 (17), 1-28.
Ratcliff, D. (n.d). 15 methods of data analysis in qualitative research, 1-6.
Smelser and Baltes. (2001). Analytic induction. International encyclopedia of the social and behavioral sciences, 1-19.
Thomas, D. (2006). A general inductive approach for analyzing qualitative evaluation data. American journal of evaluation, 2 (27),1-