INFT2150 Business Analysis

The key four steps involved in qualitative data analysis are highlighted below.

• Raw Data Management – This involves the task of data cleaning whereby the raw data is organized so as to be suitable for further data analysis. This may involve tasks such as converting the recorded audio into transcript and transforming the images in the form of videos, charts etc. for the purpose of further analysis.
• Data Reduction – The cleaned data needs to be grouped in accordance with requisite classification scheme. For this to happen, the data needs to be read a lot of times so as to identify the various patterns. Occasionally, notes would also be taken so as to facilitate aggregation of data, thereby reducing the data into more useable form. Also, the process of winnowing takes place at this stage whereby a set of codes are developed which tend to have similar meanings. Additionally general categories are also created for data reduction. Further, after the preliminary codes are prepared, there is further refining of these and from the chunks of data various clusters are obtained. Also, the codes are further refined. Thus, in this step the coding progression takes place from one level to other thereby leading to the final codes.
• Data Interpretation and themes – Various themes are identified based on the reduced grouped data with various codes. The labeling of the various clusters is done which further paves way for the labeling of specific themes. These themes are further interpreted in order to determine answers for the research questions.
• Data Representation – The interpretation and analysis of the data is carried out in order to highlight a broad thematic analysis which then develops into the narrative of the research.

In case of qualitative data analysis, the raw data collected is usually non-numerical and therefore requires preparation before interpretation of the data can be done. As a result, the concepts of coding and categorization tend to be highly relevant in qualitative data analysis. Coding refers to the broad process through which appropriate code or label is attached to a given aspect of data. There are a number of coding techniques which are available and the suitable techniques ought to be selected on the basis of the underlying research objectives coupled with the given scenario. Coding plays a crucial role in qualitative data analysis since it acts as a tool to reduce the data in a manner that significant ideas are captured. Further, the process of coding tends to aid in better understanding of phenomena under study. Coding further paves way for categorization and formation of themes which lead to answers for the research questions.

Categorization refers to the process whereby the codes tend to be labeled into various categories based on the underlying frequency and features which essentially lead to interpretation and drawing of underlying themes. Once coding is done, it is imperative to categorize the coded data into various categories based on homogeneous characteristics. This aids in building of broad themes by enabling the recognition of patterns in categorized data. Therefore, this plays a critical role in the qualitative data analysis along with coding. It 

One of the key concerns in any research is in relation to validity and reliability because the absence of these tends to hinder in the generalization of the research results in the form of a theory. The various measures that can be taken in order to ensure that the analysis is reliable and valid are as follows.

• It is essential that as part of qualitative data analysis, the background of the research needs to be highlighted which facilitate comparability and ensure that reliability is ensured. Repeating the analysis in the different context may impair reliability.
• To enhance the validity of research, triangulation of methods is recommended whereby the data is collected using various methods thereby ensuring that that the results obtained are similar and cover various aspects of the given phenomena.
• Further, the data reduction and coding needs to be checked multiple times at different stages so that the relevant aspects are highlighted in a reliable manner. Besides, different coding and classification techniques may be used to enhance the reliability of the data analysis process.
• Additionally, the researchers who are conducting the data analysis should be well trained so as to minimize researcher bias especially in the initial two stages when the raw data is reduced to a more usable form which finally is used to make conclusions. Lack of adequate training may lead to bias especially when the researcher has a particular notion about the phenomena under study.
• During the interpretation stage, it is imperative that the researcher does not adhere to normative approach but rather outlines the phenomena without being judgmental about the same thereby ensuring higher validity.  
• The coding technique used should be aligned with the underlying research questions.