Improving University Students’ Data Analysis Outputs through Effective Data Collection, Cleaning, Screening and Normalisation.

Abstract

Effective data analysis reflects an improved approach to data collection, cleaning, and screening. However, very few studies reported information on the techniques used in cleaning and screening their collected data, leading to questionable final results and interpretations, especially among university students. To address this issue, the current study examines the rigorous data collection, cleaning and screening processes for data normalization among university students in Nigeria. Using a multi-stage research methodology, 367 adapted survey instrument items were administered using Snowballing. Finally, 365 were retrieved from the respondents. Missing data were all replaced using the Series Mean (SMEAN), and outliers were appropriately addressed through z-score values and chi-square criteria. Descriptive statistical measures were used to examine the dataset and presented in several tables, a histogram, a scatterplot and a normal probability plot. The collected, cleaned, and screened data were found to have a normal distribution, helping analyze and understand the parametric data distribution, variation, and normalization. The findings provide valuable information for university students, academics, policymakers, and practitioners to adopt during data collection, cleaning and screening exercises. It was recommended that university students, lecturers, researchers, and research institutions prioritize thorough data collection, embrace transparent data cleaning, screening and reporting practices, and adopt standardized procedures to enhance data accuracy, reliability, and normalization for better data analysis and interpretation of research findings.