• No results found

Data analysis began when the researcher transforms the collected data into codable units for analysis. During this process, the real names of the participants were not visible during data analysis. Each participant received a unique code to protect her identity. After the researcher transformed all data into codes, the process of data analysis commenced. Nvivo software, which is a software used in qualitative studies, aided in the analysis of the open-ended responses (Bazeley & Jackson, 2013). Using computer software to

process qualitative data is a common practice, and is particularly helpful in processing large quantity of data (Gibbs, 2015). The researcher loaded all data into the Nvivo software for the organization and storage of data. The software allowed the researcher to store all data in one area, giving the researcher easy access to all available data. The software also helped in the organization of data because the researcher could highlight certain texts, compare the responses of one participant to another, compile texts that are thematically similar, and organize data based on emergent themes. The researcher

performed the data analysis, using the Nvivo software to improve the proper and accurate handling of data.

The researcher analyzed data using a constant comparative analysis method, which is consistent with a case study methodology. The researcher determined emergent themes from the data through the use of codes (Kolb, 2012). The first step in the constant comparative method is to reduce excess data (Kolb, 2012). Data reduction involves selection, simplification, abstraction, and transformation of raw data gathered (Kolb, 2012).

The next step after reduction was to code the data. Hewitt-Taylor (2001) proposed that in the coding process, a code will be attributed to sentences, paragraphs or sections while the researcher reads the documents of the data gathered. In order to gain consistency, each code must have a definition and an abbreviation (Kolb, 2012). For this study, codes were generated from the data instead of pre-conceived since pre-determined outcomes are not present based on literature. These codes were assigned to the data gathered. Strauss and Corbin (2008) mentioned three phases of coding: (a) open coding, (b) axial coding, and (c) selective coding. Open coding involves comparison of data and filtering out information that are clear and unclear; the researcher then tags codes for relevant information and determine emerging categories (Strauss & Corbin, 2008). In the next phase, the researcher combined the data. This was where related emerging categories were noted and grouped together (Kolb, 2012). Sub-categories were further determined through this phase (Strauss & Corbin, 2008). The final phase involved identification and selection of the core categories and these core categories were systematically connected to related core categories.

After the initial draft codes were developed, the researcher had two additional coders used those codes on the raw data. Independently, these coders coded the data, after which the researcher and the additional coders met to discuss problems with applying codes, code definitions, and inclusion/ exclusion criteria and to evaluate intercoder reliability. The researcher used Cohen’s kappa (Cohen, 1960), which prevents the inflation of reliability scores by correcting for chance agreement. The kappa measure can range from 1 to negative values no less than –1, with 1 signaling perfect agreement and 0 indicating agreement no better than chance. The intercoder reliability was .085, indicated a near perfect category of reliability.

Core themes, and their relationships to each other, were then related to each research question (Hewitt-Taylor, 2001). This process will ultimately led to an explanation of the central research problem.

III.8 Assumptions, Limitations, and Delimitations

1. The researcher assumed that participants were honest about their responses to the interview questions. The researcher emphasized confidentiality of all the information divulged.

2. The researcher assumed that Social Role Theory (Eagly, 1987; 2012) will provide a theoretical foundation in understanding the relationship between gender dynamics and women’s choices for certain business arenas. . Certain limitations and delimitations were also identified for the study. The limitations were as follows:

1. For the purposes of career choices by women, the researcher based the results of the analysis on the actual words and ideas that the participants used in the interviews. As such, the results did not identify terms as used in academic and research literature; rather, results were versed in the vernacular or using ordinary terms. Provided this limitation, the researcher may not be able to directly link the results of the study to theory.

2. Due to the use of a small sample size, as well as the use of convenience sampling, the results may not be generalized to all arenas of business or all businesswomen as those represented in the interviews. Case studies, however, do not aim at statistical generalization where results are generalizable to populations; rather, they represent an example of analytic generalization (Yin, 2009).

The following delimitations are identified:

1. This study was delimited to first generation female entrepreneurs and those in the corporate arena who have 5 to 10 years of experience.

2. This study was delimited to the experiences of the subjects. No experiments were done in acquiring data. Methods of the research were gathering data through interviews and analyzing through comparative case study.

Related documents