The Design of Human-Centered Artificial Intelligence for the Workplace by Unknown

Author:Unknown
Language: eng
Format: epub, pdf
ISBN: 9783031835124
Publisher: Springer Nature Switzerland
Published: 2025-02-15T00:00:00+00:00

5 Cross-Validation of Interaction Concept and User Involvement Models

5.1 Validation Method

In this section, we compare the interaction concept (i.e., the identified user tasks and the information architecture of the UI mockup) with three user involvement models, namely, Endsley’s Situation Awareness model (Endsley 1995), Endsley and Kaber’s Level of Automation Taxonomy (Endsley and Kaber 1999), and Rasmussen’s ladder of abstraction (Rasmussen 1974).

A crucial question arises: Are we merely comparing the user involvement models with the interaction concept, or are we validating one against the other?

A model, defined as “an abstraction of a system or of the real world concentrating on specific structural or behavioral properties and representing them in a syntactically and semantically defined language” (Aßmann et al. 2021, p. 107–108), is typically created through selection and abstraction from reality based on objective observations. According to Popper, “it must be possible for an empirical scientific system to be refuted by experience” (Popper 2005, p. 18). Therefore, we can approach scientific truth by falsifying abstracted principles (in our case, the user involvement models) against the gathered data (the interaction concept). This implies that we can validate the user involvement models against the interaction concept, posing the guiding question: Which elements of the interaction concept are missing in the model? We refer to this validation direction as “validation of the model.”

Can we also validate the interaction concept against the model? We can identify data motivated by the respective models that the interaction concept does not contain. The guiding question here is: Which elements of the model are not reflected in the interaction concept? This validation direction is termed “validation of the interaction concept.” Both validations are summarized as “cross-validation” in this section.

To perform the cross-validation, we first map the model elements and the user tasks and views of the interaction architecture to each other. Subsequently, we conduct a model validation and a validation of data completeness.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.