The Necessity of Exploratory Data Analysis How are preprocessing activities beneficial to Data Analysts and Professional Researchers in Academia

Abstract

Data analysis is used in all academic disciplines. Still, research has shown that some studies can appear ambiguous when the analyzed data needs to be sufficiently illustrated to identify trends, patterns, and other assumptions. These assumptions typically enable researchers to present the statistical summary using pertinent and self-explanatory graphical representations. Analysts want to use a method that will assist them in condensing the dataset's critical characteristics for straightforward interpretation and presentation to the audience. In addition to presenting the impact of preprocessing activities in assuring an error-free dataset before actual analysis is done, this study uncovers the trick to conducting an adequate investigation on the dataset to have a clean dataset for accurate analysis interpretation. The most popular preprocessing procedures, including missing values, outliers, and variable transformation, are listed. The study used a descriptive survey design technique and focused on using a questionnaire instrument to gather data from respondents using a Google Forms App. The information was collected using criteria such as gender, amount of data analysis experiences, institution type, and roles within the academic community. Both face validity and construct validity methods were used to validate the instrument. Chrobach's Alpha yielded a dependability index of 0.88, indicating good reliability. Since the data was prepared and collected online using Google Forms, the data collecting and collation process only took four days. Software for appropriate visualization was used for the analysis. The results demonstrated that thoroughly exploring the data and removing any bias or outliers is the first step that any data analyst must take before starting a proper analysis process. The usage of some of the tools available for cleaning datasets was also outlined, and it was recommended that amateur analysts take the time to learn how to utilize them. Before beginning their final year projects, final-year undergraduate and postgraduate students should be exposed to all exploratory data analysis methods.