Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier



Open Access Dissertation

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program


Year Degree Awarded


Month Degree Awarded


First Advisor


Second Advisor


Third Advisor


Fourth Advisor


Subject Categories

Business Administration, Management, and Operations | Experimental Analysis of Behavior | Management Information Systems


The study of individual, affect-related consequences from technology adoption and use is gaining traction in the information systems (IS) discipline. Efforts to explore affective reactions to technology have considered various positive, affective constructs (e.g., enjoyment, computer playfulness, and flow), with a more recent focus on the dark side of technology use and constructs such as technostress, technophobia, and computer anxiety. While some research has examined these negative affective responses to technology, construct definitions and relationships are not well-defined or theoretically grounded. A recent theoretical advance in IS, the Affective Response Model (ARM) categorizes affective responses to technology based on five dimensions. This three-paper dissertation explores negative affective responses to technology by (1) synthesizing the IS literature through the application of ARM, (2) proposing new affective concepts, and (3) theorizing about and testing the relationships between relevant antecedents and outcomes of these affective responses.

In paper one, an integrative literature review is conducted on computer anxiety, technophobia and technostress, the main negative affective concepts in the IS literature. The known antecedents, dimensions, and outcomes of each concept are organized into nomological networks. These nomological networks are then combined to identify inconsistencies and omissions in the literature. Further, the ARM taxonomy is applied to differentiate the three constructs and to introduce technology-induced state anxiety (TISA), a new temporal (state-like) negative response to a specific instance of technology. Two empirical studies are conducted using existing and newly developed scales, and demonstrate that computer anxiety, technophobia, technostress and TISA are conceptually and empirically distinct, laying a foundation for further exploration of how these constructs are related.

In paper two, much of the integrated nomological network from paper one is tested in the context of a laboratory experiment with a spreadsheet application. The relationship between computer anxiety, technostress and TISA is explored in more depth with the mediating influence of technostress on TISA proposed and confirmed. ARM is further extended in two ways (1) by demonstrating the impact of the characteristics of the task/organizational context, a new category of antecedents identified from paper one, and (2) connecting affective responses to computing performance outcomes (e.g. satisfaction with performance, expected future performance, and an objective measure of task accuracy). Finally, this paper concludes by evaluating how the relationship between antecedents, affective responses and performance outcomes may change with system experience. The laboratory experiment is repeated after six weeks of regular system usage to test whether the strong influence of TISA observed at time 1 diminishes as expected.

In paper 3, the research model from paper 2 is expanded by integrating positive affective concepts. It is known that positive and negative concepts are distinct and individuals can experience high levels of both positive and negative affect at the same time. Therefore, ARM is further extended by demonstrating the practical and theoretical importance of considering both positive and negative affective responses. This paper explores the domain of a less structured creative task, employing a laboratory experiment in which participants design a flyer. Computer anxiety, technostress and TISA are measured alongside enjoyment, and two newly proposed concepts, technomancy and computer enthusiasm. The unique impact of these positive and negative affective responses on performance outcomes is demonstrated. Lastly, the intervention effect of a positive mood is evaluated experimentally. Participants in a positive mood prior to working on the design task experienced more enjoyment. Those assigned a more difficult task and a less usable technology also experienced less TISA due to being in positive mood state. Positive mood also had a helpful indirect effect on performance outcomes.

The findings from the three dissertation papers have important theoretical and practical implications. A major IS theoretical framework is meaningfully applied to negative affective concepts and extended. Second, this work offers more detailed explanation of what antecedents influence certain affective concepts more, building on the omnibus and reciprocal propositions in ARM. Third, this work formally connects affective responses to computing performance outcomes. Lastly, the added benefit of considering positive concepts side-by-side with negative concepts is demonstrated. Focusing on the dark side alone is both theoretically incomplete and practically misleading.

There are also important implications for practitioners. It is shown that minimizing TISA is especially critical in the early stages of using a system, as TISA is the affective concept driving performance outcomes the most at that time. This idea holds true for both structured computing tasks and less structured, creative tasks. Also, establishing a positive mood prior to engaging with the system heightens the enjoyment experienced and reduces TISA under very challenging situations, for instance when the technology is less usable and task requirements are high. This finding confirms that a positive mood can be a positive balancing force to negative affect, indirectly preserving performance outcomes. Finally, the concluding chapter of this dissertation discusses several future research directions that build on this work.