Dissertation/Thesis Abstract

The data narrativist: Leading analytics organizations in value creation using data driven communication
by Bradbury, Judd D., Ph.D., The University of Texas at Dallas, 2016, 164; 10113850
Abstract (Summary)

This dissertation defines a new executive role of The Data Narrativist to address the emerging leadership needs of analytics organizations practicing data science. The Data Narrativist is designed to be an executive storyteller with expertise in the humanization of data driven communication. The Data Narrative Process provides an approach for managing exploratory analysis, analytics research, story creation and execution of data narratives. A new profile for this role is defined by a review of existing profiles from similar executive roles, a definition of emerging administrative roles, and an assessment of complementary methodologies.

This thesis proceeds with an investigation into the structure and techniques of documentary film applied to the practice of creating narrative data visualization in video. Structural elements in documentary video are identified to engage a broad audience not formally trained in data visualization. An empirical analysis is performed using matched pairs of narrative visualization videos highlighting particular documentary techniques. With the accumulation of user preferences, voice over audio and the participatory mode are recommended to improve audience engagement of the analytical content. This work provides guidance for the construction of effective documentary narrative visualization videos as a form of enhanced data narrative. Documentary Data Narrative is incorporated into an evaluation of three unique types of enhanced data narrative to determine their effectiveness in knowledge transfer. Two examples of enhanced data narrative using data visualization are defined as Visual Data Narrative and Documentary Data Narrative. Another is defined as Computer Generated Text Data Narrative, constructed using Natural Language Generation (NLG) technology from the field of artificial intelligence. Together the three defined forms of enhanced data narrative are evaluated along with a traditional Human Generated Text Data Narrative. Each of the four types of data narrative is constructed from a common set of quantitative data measures regarding an earnings announcement for an established technology company. Based on the results of a set of knowledge retention questions, we perform a statistical analysis to determine the knowledge transfer performance for each form of data narrative. With this analysis we make observations about the effectiveness of the four discrete forms of data narrative.

Indexing (document details)
Advisor: Schich, Maximilian
Commitee: Dess, Gregory G., Guadagno, Rosanna E., Malina, Roger F.
School: The University of Texas at Dallas
Department: Arts and Technology
School Location: United States -- Texas
Source: DAI-A 77/11(E), Dissertation Abstracts International
Source Type: DISSERTATION
Subjects: Social psychology, Management, Communication, Computer science
Keywords: Analytics, Data visualization, Executive leadership, Human computer interaction, Statistics
Publication Number: 10113850
ISBN: 9781339769073
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy
ProQuest