Estimating Resistance to Digital Technology Infusion: A Multi-Method Modeling Approach

Abstract

As technology continues to evolve, it continuously alters how work is done, how people interact, and how decisions are made. However, these transformations commonly face a critical barrier to successful implementation: resistance. Resistance arises at both the individual and organizational levels, often resulting in delays or stagnation in technology initiatives. This thesis explores how resistance might be quantitatively modeled in order to facilitate more successful technology deployment.

Building on insights from an extensive literature review, two methods are proposed that capture the common constructs present throughout the literature as well as how systems thinking can be applied. The first method leverages Large Language Models (LLMs) to perform a sentiment analysis on qualitative data, extracting sentiment related to key constructs that shape resistance attitude and behavior. The second method applies systems architecture tools and principles to analyze changes to workflows and interactions resulting from the proposed technology initiative, developing system-based proxies for resistance constructs. While both methods yield overall resistance estimates, they also generate valuable insights that can be used to strengthen change management and transition plans.

A case study within a large global corporation, featuring the deployment of the same technology across two organizations, demonstrates the application of these methods. This case study not only shows the type of insights that might be gained from utilizing these models, but exposes the limitations of each technique and the practicality for use within an enterprise. These limitations emphasize the need for large datasets and extensive system knowledge to perform a thorough analysis of the anticipated resistance.

Ultimately, this research highlights the complex technical, social, and psychological dynamics of resistance and the challenges of capturing these factors within a quantitative model. The proposed methods build upon theoretical models for change resistance and technology adoption 3 to demonstrate the value in leveraging systems thinking and cutting edge technology to generate unique resistance insights.