Title: Consumer sentiments in automotive purchases before and after COVID-19: a text-mining study

Authors: Ashok Bhattarai; Jiaxi Luo; Shih Yung Chou; Charles Ramser

Addresses: Dillard College of Business Administration, Midwestern State University, 3410 Taft Blvd., Wichita Falls, TX 76308, USA ' Dillard College of Business Administration, Midwestern State University, 3410 Taft Blvd., Wichita Falls, TX 76308, USA ' College of Arts, Sciences, Business, and Education, Winston-Salem State University, 601 S. Martin Luther King, Jr. Drive, Winston-Salem, NC 27110, USA ' Dillard College of Business Administration, Midwestern State University, 3410 Taft Blvd., Wichita Falls, TX 76308, USA

Abstract: The COVID-19 pandemic has led to shortages in the automotive industry due to a limited supply of semiconductor chips, which has created a nonlinear dynamic and chaotic business environment in the industry. This leads to the following important yet unanswered questions: 1) Is there a divergence in consumer emphases placed on the car buying process prior to and after COVID-19?; 2) How do consumer sentiment patterns affect their ratings of car dealerships prior to and after COVID-19? To answer these questions, we utilise a text-mining approach and perform an ordered probit regression analysis. Results illustrate the following. First, the sentiment keyword 'fast' had a positive impact on consumer online ratings after COVID-19, whereas 'clean' had a positive impact on consumer online ratings before COVID-19. Third, the sentiment keyword 'wait' had a negative impact on consumer online ratings after COVID-19. Fourth, the sentiment keyword 'willing' had a negative impact on consumer online ratings both before and after COVID-19. Finally, the sentiment keyword 'mess' had a negative impact on consumer online ratings both before and after COVID-19.

Keywords: automotive purchase; COVID-19; consumer sentiments; text mining.

DOI: 10.1504/IJBE.2025.143093

International Journal of Business Environment, 2025 Vol.16 No.1, pp.62 - 76

Received: 12 Jan 2023
Accepted: 12 Sep 2023
Published online: 03 Dec 2024 *