Scholarlink

Data dimension reduction techniques enable researchers to reduce the number of variables (dimensions) in a dataset while preserving essential information. In this chapter, we demonstrate the application of two widely used linear methods, Exploratory Factor Analysis (EFA) and Principal Component Analysis (PCA), for data dimensionality reduction. Both EFA and PCA aim to simplify the structure of observed data and to identify patterns of association among variables, making them particularly valuable for assessing questionnaire scale dimensionality and exploring item relationships. This chapter applies these methods to identify key service quality factors in the banking sector, with a specific focus on public sector banks in Kanpur City, Uttar Pradesh, India. A survey of 506 participants was conducted in 27 bank branches using an 18 item service quality questionnaire. Our methods and results offer actionable insights for deriving latent customer service quality constructs for Indian public sector banks. The analysis in this chapter mainly aims to provide practical guidance for implementing EFA and PCA in both R and Python, including the use of polychoric correlations to appropriately handle Likert-scale responses. More broadly, the chapter illustrates the applicability of EFA and PCA, which can be applied across domains such as public policy, social science and management research.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dimensionality Reduction Techniques: Application to Indian Bank Customer Survey Data

  • 章节

  • 发表日期：

    2026年

查看全文
  • Maharnab Naha,
  • Kanish Debnath,
  • Karan Singh Khati

摘要

Data dimension reduction techniques enable researchers to reduce the number of variables (dimensions) in a dataset while preserving essential information. In this chapter, we demonstrate the application of two widely used linear methods, Exploratory Factor Analysis (EFA) and Principal Component Analysis (PCA), for data dimensionality reduction. Both EFA and PCA aim to simplify the structure of observed data and to identify patterns of association among variables, making them particularly valuable for assessing questionnaire scale dimensionality and exploring item relationships. This chapter applies these methods to identify key service quality factors in the banking sector, with a specific focus on public sector banks in Kanpur City, Uttar Pradesh, India. A survey of 506 participants was conducted in 27 bank branches using an 18 item service quality questionnaire. Our methods and results offer actionable insights for deriving latent customer service quality constructs for Indian public sector banks. The analysis in this chapter mainly aims to provide practical guidance for implementing EFA and PCA in both R and Python, including the use of polychoric correlations to appropriately handle Likert-scale responses. More broadly, the chapter illustrates the applicability of EFA and PCA, which can be applied across domains such as public policy, social science and management research.