![3
Higher education decision makers are keen to use the vast and
still-
growing volume of data on students, faculty, staff, and insti-
tutions themselves. More data, it may be reasoned,
will produce better
decisions. On the surface that can be true, and yet the larger volume of
data does not necessarily ensure better decision making. Along with
more data comes the need to use contextualized knowledge of the higher
education organ
ization and analytics strategies that account for the
unique situation or population
under study, and every
one must be mind-
ful of privacy, ethical, and overall responsible use of the data. While the
allure of vast quantities of data offer the possibility of greater student
success and more effectively managed institutions, higher education
leaders must consider how data analytics can be harnessed success-
fully, how strategies for good data governance and orga
nizational
strategies can support informed decision making, and how and where
issues of privacy and security must be addressed.
In the article entitled “Data to Knowledge to Results: Building an
Analytics Capability,” Davenport et al. (2001) foresaw the impact of the
data tsunami on orga
nizational decision making and lamented that “in
the rush to use computers for all transactions, most organ
izations have
neglected the most impor
tant step in the data transformation pro
cess:
the
human realm of analyzing and interpreting data and then acting on
[ 1 ]
Data Analytics and the Imperatives
for Data-
Informed Decision Making in
Higher Education
Karen L. Webber and Henry Y. Zheng](https://image.slidesharecdn.com/19124515-250516171438-325039c7/85/Big-Data-On-Campus-Data-Analytics-And-Decision-Making-In-Higher-Education-Karen-L-Webber-20-320.jpg)
![30
Gosh,
you’ve
really got some nice toys
here.
—Roy Batty, Blade Runner (1982), set in 2019
In the summer of 1956, luminaries of mathe
matics and information
sciences gathered at Dartmouth College for two months to hold in-
depth discussions about what the group or
ga
nizer John McCarthy
termed artificial intelligence. McCarthy’s proposal to the Rocke
fel
ler
Foundation to support the summer meeting provocatively and pre-
sciently asserted: “If a machine can do a job, then an automatic calcula-
tor can be programmed to simulate the machine. The speeds and mem-
ory capacities of pre
sent computers may be insufficient to simulate
many of the higher functions of the
human brain, but the major ob-
stacle is not lack of machine capacity, but our inability to write pro-
grams taking full advantage of what we have” (McCarthy et al. 1955).
Participants in this summer meeting included, among
others, Arthur
Samuel, who would
later in the de
cade coin the term machine learning
as he developed a computer program that could win a checkers game
against a
human being; Ray Solomonoff, who developed algorithmic
information theory that machine learning is probabilistic and can be
trained on existing data to solve new prob
lems; Marvin Minsky, who
would go on to co-
found the AI Lab at MIT; and Claude Shannon at
[ 2 ]
Big Data and the Transformation of
Decision Making in Higher Education
Braden J. Hosch](https://image.slidesharecdn.com/19124515-250516171438-325039c7/85/Big-Data-On-Campus-Data-Analytics-And-Decision-Making-In-Higher-Education-Karen-L-Webber-47-320.jpg)
![Half an hour more went by; and then was heard the sound of many
feet passing along through some chamber near. At the end of above
five minutes the door opened, and Monsieur de Tronson led in an
elderly lady habited as if for a journey.
"Madame de Langdale," said the secretary of the cabinet, addressing
Lucette, "Madame de Lagny, with whom you passed last night, will
have the pleasure of accompanying you and Monsieur de Langdale
on your journey. The carriage has been ready for an hour; but, the
council having sat later than usual, I could not leave my post.
Monsieur will do me the honor of accompanying me to his chamber
below, where I will put him in possession of his money and his safe-
conduct, together with his baggage, while you prepare for travelling,
which, as it is, must, I fear, be protracted into the night."
Edward followed him down several flights of steps, conversing with
him, as he went, upon the arrangements for their journey, telling
him that he feared from his servant's information they would be
obliged to proceed beyond Niort to St. Martin des Rivières, and that,
consequently, at least two days more than he had calculated upon
must pass ere he could fulfil the promise he had given to return.
But De Tronson seemed thoughtful and absent; for, in truth, he had
just come from a painful scene;[3] and, although he heard, and
answered all his young companion said, it was by an effort, and
evidently without interest.
All the arrangements were soon made, however. Edward's property
was restored to him; the tradesmen he and Lucette had employed
were paid; and then the secretary led him to the little court, where
stood one of the large clumsy carriages of the day with four tall
horses. A stout man on horseback was also there, holding by the
rein the horse which Jacques Beaupré had ridden to Nantes, and, as
no beast had been provided for Pierrot, he mounted beside the
coachman. Lucette and her companion were already in the vehicle,
and, with a kind adieu from M. de Tronson, Edward took his place
beside them, and the vehicle rolled on.](https://image.slidesharecdn.com/19124515-250516171438-325039c7/85/Big-Data-On-Campus-Data-Analytics-And-Decision-Making-In-Higher-Education-Karen-L-Webber-56-320.jpg)
![Lucette's heart was asking her the same question; but by this time
the Duc de Rohan was advancing to meet her and her companion,
and in a moment more he was near enough to take Madame de
Lagny's hand and raise it courteously to his lips.
"You have come to a rude place, madame," he said, "and among
somewhat rude men; but we must do what we can to make your
stay tolerable."
"Oh, my lord duke," replied the lady, with a courtly inclination of the
head, "I must away as soon as possible. I am expected back at the
court directly. But where is Monsieur de Langdale? I do not see him."
"He is in the chateau, madame," replied the duke; "but he has been
telling me so strange a tale that I have judged it best, before he and
this—["girl," he was in the act of saying; but he checked himself, and
substituted the words "young lady"]—before he and this young lady
meet again, to have from her lips and from yours what are the facts
of the case. Pray, let us go in."
"The facts of the case are very simple, my lord," replied the old lady,
with some stiffness. "Monsieur de Langdale is the husband of this
young lady, formerly Mademoiselle de Mirepoix, whom you do not
seem to recognise, my lord duke, though she is your near of kin. He
married her in the presence of the cardinal and the whole court."
"More impudent varlet he!" exclaimed the duke, angrily. "And you,
mademoiselle,—what have you to say to all this fine affair? Why, you
are a mere child! This marriage can never stand!—without any one's
consent! It is a folly!"
"Not at all, duke," said Madame de Lagny. "Pray, recollect, sir, that
Madame de Rambouillet was married at twelve,—I myself at sixteen.
Madame is nearly fifteen, she tells me; and, as to the marriage not
standing, you will find yourself much mistaken. The man who made
it is not one to leave any thing he undertakes incomplete, as you will
discover. They are as firmly married as any couple in the land, and
that with the full authority of the king, which in this realm of France](https://image.slidesharecdn.com/19124515-250516171438-325039c7/85/Big-Data-On-Campus-Data-Analytics-And-Decision-Making-In-Higher-Education-Karen-L-Webber-60-320.jpg)
![wretched young man, all bleeding, leaning his wounded head upon
his mother's bosom.
The executioner had missed his stroke. Again and again he missed
it. He complained of the sword: a heavier one was handed up to
him; but still his shaking arm refused to perform its hideous office,
till, after more than thirty blows,[4] the head of the unhappy young
man was literally hacked off, almost at his mother's feet.
The noble woman raised her hands and her eyes to heaven,
exclaiming, "I thank thee, O God, that my son has died a martyr and
not a criminal!"
The last acts of the terrible drama Edward did not see. He felt as if
his heart would burst with the mingled feelings of indignation and
horror which all he had beheld awakened; and after the second or
third blow he kept his eyes resolutely bent down, till the pressure of
the crowd relaxed as the spectators of the bloody scene began to
disperse. Then, sick at heart, and with a strange feeling of hatred for
the world, he turned his steps back to the inn. He was in no mood
for conversation with any one.](https://image.slidesharecdn.com/19124515-250516171438-325039c7/85/Big-Data-On-Campus-Data-Analytics-And-Decision-Making-In-Higher-Education-Karen-L-Webber-78-320.jpg)
Big Data On Campus Data Analytics And Decision Making In Higher Education Karen L Webber
- 1. Big Data On Campus Data Analytics And Decision Making In Higher Education Karen L Webber download https://ebookbell.com/product/big-data-on-campus-data-analytics- and-decision-making-in-higher-education-karen-l-webber-38249030 Explore and download more ebooks at ebookbell.com
- 2. Here are some recommended products that we believe you will be interested in. You can click the link to download. Big Data On Kubernetes Neylson Crepalde Thariq Mahmood https://ebookbell.com/product/big-data-on-kubernetes-neylson-crepalde- thariq-mahmood-58491926 Big Data On Kubernetes Neylson Crepalde https://ebookbell.com/product/big-data-on-kubernetes-neylson- crepalde-74148298 Pocket Data Mining Big Data On Small Devices 1st Edition Mohamed Medhat Gaber https://ebookbell.com/product/pocket-data-mining-big-data-on-small- devices-1st-edition-mohamed-medhat-gaber-4408678 The Intelligent Marketers Guide To Data Privacy The Impact Of Big Data On Customer Trust 1st Ed Robert W Palmatier https://ebookbell.com/product/the-intelligent-marketers-guide-to-data- privacy-the-impact-of-big-data-on-customer-trust-1st-ed-robert-w- palmatier-9962180
- 3. Big Data Analysis On Global Community Formation And Isolation Sustainability And Flow Of Commodities Money And Humans 1st Ed 2021 Ikeda https://ebookbell.com/product/big-data-analysis-on-global-community- formation-and-isolation-sustainability-and-flow-of-commodities-money- and-humans-1st-ed-2021-ikeda-32623574 2021 International Conference On Big Data Analytics For Cyberphysical System In Smart City Volume 2 Lecture Notes On Data Engineering And Communications Technologies 103 1st Ed 2022 Mohammed Atiquzzaman https://ebookbell.com/product/2021-international-conference-on-big- data-analytics-for-cyberphysical-system-in-smart-city- volume-2-lecture-notes-on-data-engineering-and-communications- technologies-103-1st-ed-2022-mohammed-atiquzzaman-38473954 Processing Big Data With Azure Hdinsight Building Realworld Big Data Systems On Azure Hdinsight Using The Hadoop Ecosystem Yadav https://ebookbell.com/product/processing-big-data-with-azure- hdinsight-building-realworld-big-data-systems-on-azure-hdinsight- using-the-hadoop-ecosystem-yadav-22022414 The 2021 International Conference On Machine Learning And Big Data Analytics For Iot Security And Privacy Spiot2021 Volume 1 Lecture Notes On Data Engineering And Communications Technologies 97 1st Ed 2022 John Macintyre Editor https://ebookbell.com/product/the-2021-international-conference-on- machine-learning-and-big-data-analytics-for-iot-security-and-privacy- spiot2021-volume-1-lecture-notes-on-data-engineering-and- communications-technologies-97-1st-ed-2022-john-macintyre- editor-36088748 The 2021 International Conference On Machine Learning And Big Data Analytics For Iot Security And Privacy Spiot2021 Volume 2 Lecture Notes On Data Engineering And Communications Technologies 98 1st Ed 2022 John Macintyre Editor https://ebookbell.com/product/the-2021-international-conference-on- machine-learning-and-big-data-analytics-for-iot-security-and-privacy- spiot2021-volume-2-lecture-notes-on-data-engineering-and- communications-technologies-98-1st-ed-2022-john-macintyre- editor-38210404
- 6. BIG DATA ON CAMPUS
- 7. This page intentionally left blank
- 8. BIG DATA ON CAMPUS Data Analytics and Decision Making in Higher Education Johns Hopkins University Press | Baltimore EDITED BY KAREN L. WEBBER AND HENRY Y. ZHENG
- 9. © 2020 Johns Hopkins University Press All rights reserved. Published 2020 Printed in the United States of Amer i ca on acid- free paper 9 8 7 6 5 4 3 2 1 Johns Hopkins University Press 2715 North Charles Street Baltimore, Mary land 21218-4363 www.press.jhu.edu Library of Congress Cataloging- in- Publication Data Names: Webber, Karen L., editor. | Zheng, Henry Y., editor Title: Big data on campus : data analytics and decision making in higher education / edited by Karen L. Webber and Henry Y. Zheng. Description: Baltimore : Johns Hopkins University Press, 2020. | Includes bibliographical references and index. Identifiers: LCCN 2019059971 | ISBN 9781421439037 (paperback) | ISBN 9781421439044 (ebook) Subjects: LCSH: Universities and colleges— Administration— Data pro cessing. Classification: LCC LB2341.B4785 2020 | DDC 378.1/01— dc23 LC rec ord available at https://lccn.loc.gov/2019059971 A cata log rec ord for this book is available from the British Library. Special discounts are available for bulk purchases of this book. For more informa- tion, please contact Special Sales at specialsales@press . jhu . edu. Johns Hopkins University Press uses environmentally friendly book materials, including recycled text paper that is composed of at least 30 percent post- consumer waste, whenever possible.
- 10. Foreword, by Christine M. Keller vii Acknowledgments xi PART I. Technology, Digitization, Big Data, and Analytics Maturity as the Enabling Conditions for Data- Informed Decision Making 1 Data Analytics and the Imperatives for Data- Informed Decision Making in Higher Education 3 Karen L. Webber and Henry Y. Zheng 2 Big Data and the Transformation of Decision Making in Higher Education 30 Braden J. Hosch 3 Predictive Analytics and Its Uses in Higher Education 50 Henry Y. Zheng and Ying Zhou PART II. The Ethical, Cultural, and Managerial Imperatives of Data- Informed Decision Making in Higher Education 4 Limitations in Data Analytics: Potential Misuse and Misunderstanding in Data Reports and Visualizations 79 Karen L. Webber and Jillian N. Morn 5 Guiding Your Organ ization’s Data Strategy: The Roles of University Se nior Leaders and Trustees in Strategic Analytics 103 Gail B. Marsh and Rachit Thariani 6 Data Governance, Data Stewardship, and the Building of an Analytics Orga nizational Culture 122 Rana Glasgal and Valentina Nestor CONTENTS
- 11. vi Contents PART III. The Application of Analytics in Higher Education Decision Making: Case Studies 7 Data Analytics and Decision Making in Admissions and Enrollment Management 151 Tom Gutman and Brian P. Hinote 8 Predictive Analytics, Academic Advising, Early Alerts, and Student Success 177 Timothy M. Renick 9 Constituent Relationship Management and Student Engagement Lifecycle 198 Cathy A. O’Bryan, Chris Tompkins, and Carrie Hancock Marcinkevage 10 Learning Analytics for Learning Assessment: Complexities in Efficacy, Implementation, and Broad Use 228 Carrie Klein, Jaime Lester, Huzefa Rangwala, and Aditya Johri 11 Using Data Analytics to Support Institutional Financial and Operational Efficiency 260 Lindsay K. Wayt, Susan M. Menditto, J. Michael Gower, and Charles Tegen PART IV. Concluding Comments 12 Data- Informed Decision Making and the Pursuit of Analytics Maturity in Higher Education 285 Karen L. Webber and Henry Y. Zheng Contributors 311 Index 315
- 12. Amid declining enrollments, daunting resource constraints, and flagging public support, higher education leaders can still transform colleges and universities for the benefit of students and society. A renewed commit- ment to the institution- wide use of data analytics in strategic decision making has extraordinary potential to accelerate efforts to advance in- stitutional goals, improve quality and efficiency, strengthen student out- comes, and enhance teaching, learning, and advising. Data analytics adoption and use in higher education for decision mak- ing is currently stagnant and uneven— despite the rapid development of new information technologies, wider access to analytical tools, and the acceleration of the “data revolution” in other industries. In a recent survey of provosts and chief academic officers among US colleges, Inside Higher Ed (Jaschik and Lederman 2019) found that less than 20% of provosts across public and private institutions believe that their univer- sities use data very effectively to inform institution decision making. Tom Davenport et al. (2001) lamented nearly two de cades ago that “in the rush to use computers for all transactions, most organ izations have neglected the most impor tant step in the data transformation pro cess: the human realm of analyzing and interpreting data and then acting on the insights” (18). The use of analytics to assist in the translation of data into actionable insights remains a major barrier at many colleges and universities. To increase the sense of urgency for action, the leaders of three higher education associations— the Association for Institutional Research (AIR), EDUCAUSE, and National Association of College and Univer- sity Business Officers (NACUBO)— developed a joint statement in 2019 prompting higher education leaders to reenergize efforts to unleash the FOREWORD
- 13. viii Foreword power of data and analytics and support decision making for the ben- efit of students and institutions (changewithanalytics . com). Each of the three agencies strongly believes that using data to better understand stu- dents and institutional operations paves the way to developing innova- tive approaches for improved student recruiting, better student out- comes, greater institutional efficiency and cost- containment, and much more. The joint statement includes six princi ples of action to accelerate the meaningful use of analytics and take advantage of the insights de- rived from data to make the decisions and take the actions that set up higher education for a successful future. Our associations have made our own commitment to advancing and supporting the efforts of insti- tutional leaders through the sharing of targeted resources, success sto- ries, and implementation guides. Webber and Zheng’s book on data analytics in higher education re- inforces the urgency and adds a valuable resource for higher education leaders. The book describes the conceptual under pinnings of the roles of data analytics within higher education as well as recent innovations in analytic models, new types of data and their curation, and digital me- dia. Several chapters focus on the critical importance of pro cesses and structures such as a mission- focused data strategy and robust governance policies. Importantly, a strong emphasis is placed on the human factors in successfully using data analytics for decisions such as an institution- wide commitment to a culture of data use and data literacy. The book will be valuable to researchers and prac ti tion ers alike. It is written by experi- enced higher education professionals who are recognized champions of data- informed decision making and provides case studies, examples of the application of data analytics, and helpful checklists for readers to scan their own institutional environments for opportunities. AIR’s core mission is to empower higher education professionals to use data and analytics to make decisions and take actions that benefit students and institutions and improve higher education. As executive director and CEO of AIR, I believe that Webber and Zheng’s book re- flects a core purpose of AIR and the values of our members and stake- holders. Each chapter brings practical and thoughtful insights from some of the best thinkers and leaders of data analytics in higher educa-
- 14. Foreword ix tion. The book is a timely and useful resource for higher education lead- ers and stakeholders, and I am confident it will add to our ability to harness the power of data and analytics for the success of higher edu- cation institutions and students. Christine M. Keller, PhD Executive Director and CEO Association for Institutional Research References Davenport, T. H., J. G. Harris, D. W. DeLong, and A. L. Jacobson. 2001. “Data to Knowledge to Results: Building an Analytics Capability.” California Management Review 43 (2): 117–138. Jaschik, S., and D. Lederman. 2019. “The 2019 Inside Higher Education Survey of College and University Chief Academic Officers— A Study by Gallup and Inside Higher Education.” Inside Higher Education. January 23, 2019. https://www .insidehighered.com/news/survey/2019-inside-higher-ed-survey-chief-academic -officers.
- 15. This page intentionally left blank
- 16. Our sincerest appreciation is extended to Johns Hopkins University Press, particularly Editorial Director Greg Britton, who responded with encouragement to our initial inquiry about this pos si ble volume. We also acknowledge JHUP’s assistant editors Catherine Goldstead and Kyle Gipson, who patiently answered many questions throughout manuscript preparation, and the production team at the Press. To our colleagues who contributed to this volume, we extend sincere gratitude. Without you, this book would not have been pos si ble; your expertise in specific facets of data and analytics in higher education helped us develop full and meaningful information on this impor tant and rapidly changing topic. Readers of this book will have a much better understanding of how higher education researchers and administrators are thinking about data analytics, and we hope that the implications discussed in each chapter urge readers to ponder the state of data analytics today and how it might contribute to its continued impact in higher education. We thank IR colleagues who answered our email questions and of- fered insights that strengthened the discussion. Special thanks to Timo- thy Chow for his review and comments on portions of several chapters. We also thank our work colleagues at the University of Georgia’s Insti- tute of Higher Education and the Ohio State’s Office of Strategy Man- agement, who kindly offered their support and assistance. We especially thank UGA’s Institute of Higher Education doctoral student Amy Yan- dell for her assistance with manuscript preparation, IHE administrative man ag er Suzanne Graham for help with review of drafts, and the Lehigh University Office of Institutional Research and Strategic Analytics for its support and use of its experience in our writings. ACKNOWLEDGMENTS
- 17. xii Acknowledgments We also thank our families for their support. They allowed us the freedom and quiet space to endeavor the development and finalization of drafts and edits to each chapter. We are better academic profession- als because of your quiet patience and ever- present support. 349-86472_Webber_ch01_3P.indd 12 8/25/20 9:18 PM
- 18. PART I. Technology, Digitization, Big Data, and Analytics Maturity as the Enabling Conditions for Data- Informed Decision Making
- 19. This page intentionally left blank
- 20. 3 Higher education decision makers are keen to use the vast and still- growing volume of data on students, faculty, staff, and insti- tutions themselves. More data, it may be reasoned, will produce better decisions. On the surface that can be true, and yet the larger volume of data does not necessarily ensure better decision making. Along with more data comes the need to use contextualized knowledge of the higher education organ ization and analytics strategies that account for the unique situation or population under study, and every one must be mind- ful of privacy, ethical, and overall responsible use of the data. While the allure of vast quantities of data offer the possibility of greater student success and more effectively managed institutions, higher education leaders must consider how data analytics can be harnessed success- fully, how strategies for good data governance and orga nizational strategies can support informed decision making, and how and where issues of privacy and security must be addressed. In the article entitled “Data to Knowledge to Results: Building an Analytics Capability,” Davenport et al. (2001) foresaw the impact of the data tsunami on orga nizational decision making and lamented that “in the rush to use computers for all transactions, most organ izations have neglected the most impor tant step in the data transformation pro cess: the human realm of analyzing and interpreting data and then acting on [ 1 ] Data Analytics and the Imperatives for Data- Informed Decision Making in Higher Education Karen L. Webber and Henry Y. Zheng
- 21. 4 Technology, Digitization, Big Data the insights.” According to Davenport et al., companies have empha- sized impor tant technology and data infrastructures, but they have not attended to the orga nizational, cultural, and strategic changes necessary to leverage their investments. In other words, having the data but not using them to generate actionable insights to achieve better organizational outcomes was the prob lem. Eigh teen years later, that message has been heard loud and clear among organ izations across the world. Intel Corpo- ration CEO Brian Krzanich (in Gharib 2018) called data the “new oil” that is essential to orga nizational agility and survival. He further sur- mised that data and their use in analytics will have a fundamental impact on most industries across the board. Like the business community, the higher education sector is feeling similar pressures from the data analytics movement. Facing growing competition, rising education costs, and shifting demographic trends, the highly pressurized and competitive higher education environment today has shown the importance of a deep commitment to data- informed decision support (Gagliardi and Turk 2017; Swing and Ross 2016). Scholarly publications on the status and challenges of learning analyt- ics are becoming more frequent (e.g., Arnold and Pistilli 2012; Drachsler and Greller 2016; Khalil and Ebner 2015; Viberg et al. 2018) and is- sues related to data analytics have been featured prominently in recent years in EDUCAUSE’s list of top 10 information technology (IT) issues. For example, in the 2019 list (Grajek 2019), issue #3 concerns privacy, issue #6 addresses the data- enabled institution, and issue #8 speaks to data management and governance. In a recent interview, Michael Crow, president of Arizona State University (ASU) and a nationally known in- novator in higher education, commented on how data analytics in- forms decision making at ASU: For us, to be a public university means engaging the demographic complexity of our society as a whole. It means understanding that demographic complexity. It means designing the institution to deal with that demographic complexity. And it means accepting highly differenti- ated types of intelligence: analytical intelligence, emotional intelligence. Students are not of one type but are of many, many types. Taking all of
- 22. Data Analytics and Data- Informed Decision Making 5 that and overlaying it with hundreds of degree programs results in so many variables and so many dimensions of complexity that you actually can’t operate the institution unless you make a fundamental switch and say to yourself that, at the end of the day, it is just about analytics. (Bichsel 2012, 16) Despite some newfound emphasis on data analytics, most higher ed- ucation officials are not yet adept at using analytics to support institu- tional decision making. In a recent analy sis of more than 250 confer- ence papers and journal publications on learning analytics, Viberg et al. (2018) reported that although the field of learning analytics is matur- ing, there is little evidence that shows improvement of student outcomes from learning analytics or that analytics has yet to be deployed widely. Similarly, in a recent survey of provosts and chief academic officers among US colleges, Inside Higher Education analysts (Jaschik and Leder- man 2019) found that only 16% of private university provosts and 19% of public university provosts believe that their universities use data very effectively to inform campus decision making. This predicament is of- ten described as being “data rich but information poor” (Reinitz 2015), and precisely how Davenport et al. (2001) described the industry al- most 18 years ago. Clearly, for data- informed decision making to take root in higher education, we must have conceptual clarity on what de- fines data- informed decision making and how it can be practiced. This and the subsequent chapters in this book seek to explain and illustrate how data analytics can support a data- informed decision making cul- ture in higher education. While the focus of discussions in this book re- lates to data analytics that affect student success and institutional ad- ministration, we heartily acknowledge that Big Data and techniques such as predictive analyses are being used in faculty member research. The creation of new knowledge is indeed a vital endeavor, and Big Data labs and advanced computing centers with high- capacity computing are enabling researchers to investigate impor tant questions such as chang- ing weather patterns and their current and predicted impact on living conditions, food sources, and energy consumption. Data analytics has the potential to help researchers move society forward in many ways.
- 23. 6 Technology, Digitization, Big Data Further, the discussions in this book focus on data analytics in US higher education, but we fully acknowledge that similar trends and activities are happening in higher education around the world. Although the ex- amples provided herein are from US institutions, data analytics poses similar challenges and opportunities in higher education across the globe. Data- Informed Decision Making vs. Data- Driven Decision Making In higher education and other industries, the terms data-informed and data-driven are often used interchangeably in describing how data ana- lytics supports orga nizational decision making. However, these two terms carry dif fer ent meanings, and therefore it is impor tant to discuss their differences and similarities so that there is a conceptual clarity as we move on to discuss data- informed decision making in the remainder of this book. • Data- Driven Decision Making (DDDM) gained strength in the 1980s. It focuses on decision algorithms, heuristics, and decision rules that empower decision pro cesses and minimize human factors (let data speak for itself). • Data- Informed Decision Making (DIDM), more recently intro- duced, focuses on leveraging data to generate insights to provide the contexts and evidence base for formulating decisions (let us figure out what data tells us). According to Heavin and Power (2017), DDDM refers to the collec- tion and analy sis of data to make decisions. Data “drive” the decision making, and conclusions are made using verifiable data or facts. It is the practice of basing decisions on the analy sis of data rather than purely on intuition (Provost and Fawcett 2013). DDDM is a decision pro cess guided by a set of algorithms supported by both historical and current data ele ments. These algorithms can be a set of mathematical formulas, an engineering model, or a machine learning module. The decisions— typically routine and operational in nature— are supported and even
- 24. Data Analytics and Data- Informed Decision Making 7 suggested by the algorithms so that human decision makers do not need to add input; most algorithms produce decisions that are automatically accepted by the computer systems. For example, when student academic rec ords are read and pro cessed by a degree audit program, the algorithm built in to that program will evaluate the students’ eligibility for degree completion. The program can generate a set of courses that need to be taken by each student and may even suggest dif fer ent pathways for degree completion. When a student has completed all degree require- ments and is eligible for graduation, an automated procedure may alert the student to file application for graduation and for inclusion in the next commencement. While a number of articles or other written documents use the terms DIDM and DDDM interchangeably, we argue that the“drive”in DDDM implies that data determine the direction of the decision- making pro cess and decision makers typically accept the decision recommendations. Many of the decisions made in business organ izations are DDDM even though we may not even realize it. For example, Walmart stores nation- ally restock their shelves when inventory tracking systems detect low inventory and an order will be automatically placed for the suppliers to restock. In higher education, when students miss a deadline to pay fees or exceed the credit hours limit for the semester, an email will be auto- matically generated to remind the students and the system will block the students’ ability to enroll for the semester. While DDDM systems exist and can provide some advantages in ensuring some proactive prompts (when decision logic is fully implemented), we believe that data- informed decision making is more helpful and robust in most decision situations when human intelligence and flexibility are required. Therefore, the fo- cus of this book is more about DIDM and less on DDDM. DIDM recognizes that human judgment is a key ele ment in complex, dynamic, and strategic decision making. Because of the complexities, DIDM involves many more variables than a set of algorithms may be able to effectively address. Politics, human sensitivity, orga nizational val- ues, and timing considerations are just some examples of why com- puter programs cannot fully be incorporated to make “data- driven” de- cisions for many dynamic decision situations.
- 25. 8 Technology, Digitization, Big Data We define DIDM as the pro cess of organ izing data resources, con- ducting data analy sis, and developing data insights to provide the con- texts and evidence base for formulating orga nizational decisions. In DIDM, data are just the evidence base, while the decision context is very much as impor tant as, if not more impor tant than, the data alone. Higher education leaders, even when equipped with sufficient data and excel- lent analy sis, will need to draw on their professional experience, intu- ition, po liti cal acumen, ethical standards, and strategic considerations in making their decisions. Data are the impor tant part of the decision equation but not the only part that drives the decision (Knapp, Cop- land, and Winnerton 2007).According to Maycotte (2015),“Being data- informed is about striking a balance in which your expertise and under- standing of information plays as great a role in your decisions as the information itself. In the analogy of flying an airplane—no matter how sophisticated the systems onboard are, a highly trained pi lot is ultimately responsible for making decisions at critical junctures. The same is true in a business organ ization” (1). Given the recent tragic loss of two Boeing 737 Max airplanes, seemingly due to faulty control algorithms, Maycotte’s analogy is appropriate yet troubling. The Importance of Clearly Differentiating between DDDM and DIDM DIDM has its roots in the orga nizational learning theories in organiza- tional management lit er a ture (Goldring and Berends 2009; Winkler and Fyffe 2016). Orga nizational learning is the pro cess by which members of an organ ization acquire and use information to change and implement action (Beckhard 1969). Organ izations that have knowledge systems distributed across functional units and individuals as well as embedded in the culture, values, and routines of the organ izations are undergoing the pro cess of orga nizational learning. In this way, data can serve as a catalyst to propel orga nizational learning. Leaders can use data to put into place mechanisms to support individual and collective learning sur- rounding data (Pfeffer 1998). A few more comments may help examine the differences between these two forms of decision making:
- 26. Data Analytics and Data- Informed Decision Making 9 • DIDM is a more relevant and useful concept in the context of higher education because the decision context is dynamic; • DIDM acknowledges that data are not perfect, in the sense that not all data are available and not all available data are accurate; • DIDM acknowledges that analyses and algorithms are not perfect; models and algorithms are based on the information available, and human interpretation is needed; • Orga nizational decision making is more nuanced than most algorithms can predict; and • Human interactions and environmental factors are not as routine and are more likely to change. No doubt, data are invaluable and critical sources of insights for higher education organ izations today. However, data analytics alone does not drive decisions, especially those strategic and operational de- cisions that have complex and dynamic contextual factors. For exam- ple, many universities employ predictive models to help them identify and recruit students and make admissions decisions. However, these pre- dictive models do not replace the careful review and reading of the admission files and supporting documents by the admissions counsel- ors. Many intangible factors need to be accounted for in such decisions. It would be callous and arbitrary if admissions offices relied entirely on quantifiable data and decision algorithms. In order to fulfill the missions of higher education that include teach- ing and learning, research and discovery, and public and community ser vices, higher education officials engage in human interactions with constituents or stakeholders. The idea of having super- algorithms to drive decisions and actions may have some appeal in the routinized and stable decision situations such as degree audit. However, we believe that DIDM is a better paradigm and concept to embrace, particularly in stra- tegic and operational decision making pro cesses that involve human judgment, po liti cal sensitivity, and ethical considerations. For DDDM to work well, data need to be clean, stable, and consistent, and regu- larly updated. Such an ideal situation is often not available in higher education.
- 27. 10 Technology, Digitization, Big Data Many institutions, even those equipped with the best data ware houses and business intelligence systems, face many challenges in data manage- ment. Due to inconsistent data standards and definitions, varying ef- forts in data quality control, and lack of strong data governance prac- tices, it is not unusual that dif fer ent numbers are produced for a seemingly identical question. A classic example is the calculation of fac- ulty full- time equivalents (FTEs). The Offices of Institutional Research, Human Resources, Faculty Affairs, and academic departments may all be able to produce their own FTE numbers. Depending on what data definition is used, it is pos si ble that all answers are technically correct but each is derived for a dif fer ent context (Zheng 2015). Additionally, in the age of the Internet of Things (IoT),* the speed, volume, and variety of data available for decision analy sis are over- whelming and they limit decision makers’ ability to pro cess all avail- able data quickly enough to use predetermined algorithms to drive de- cisions. Chai and Shih (2017) point out that there is a growing belief that sophisticated algorithms can explore huge databases and find rela- tionships in de pen dent of any preconceived theory and hypotheses. The assumption is: The bigger the data, the more power ful and precise are the findings. However, this belief may be misguided and risky. There is high potential for more data sources and new data ele ments for which the current algorithms cannot account. Algorithms can include small bi- ases in data that may be compounded. Because many machine learning applications do not offer a transparent way to see the algorithms or logic behind recommendations (O’Neil 2016), some business leaders call for “explainable algorithms.” Despite all the hype about Big Data, data cannot be very useful unless they can be analyzed in a timely way to develop contextualized meaning (Lane and Finsel 2014). In its 2012 report “Analytics in Higher Education: Benefits, Barriers, Pro gress, and Recommendations,” EDUCAUSE formally defined ana- lytics as “the use of data, statistical analy sis, and explanatory and pre- dictive models to gain insight and act on complex issues”(Bichsel 2012, 6). *For a brief definition and discussion on IoT, see: https://www.forbes.com/sites/jacob morgan/2014/05/13/simple-explanation-internet-things-that-anyone-can-understand/#5318 c8971d09.
- 28. Data Analytics and Data- Informed Decision Making 11 Analytics programs can offer institutions a way to be responsive to the increasingly challenging demands of orga nizational per for mance and strategic development they now face. EDUCAUSE’s definition of ana- lytics is in alignment with the data- informed decision making concept. It recognizes the need for data to be statistically analyzed, explained, and used to support complex decision situations. DIDM is also impor tant to orga nizational decision making in higher education because many strategic, operational, and management deci- sions that leaders face are dynamic, complex, and more nuanced than most algorithms can predict well. The organ ization’s unique and nu- anced issues make it difficult to suggest a perfect decision. According to a McKinsey survey of US companies (Marr 2018), only 18% of busi- ness leaders believe they can gather and use data insights effectively. Concerns include the need for proper analy sis, how data are communi- cated to decision makers, who, in turn, act from the insights. This find- ing is similar to what we discussed earlier in this chapter: higher educa- tion leaders’ perception that a small percentage of provosts and chief academic officers believe that their universities use data very effectively to inform campus decision making (Jaschik and Lederman 2019). Enabling Conditions for DIDM in Higher Education Institutions DIDM in higher education does not happen overnight, nor, in most cases, smoothly. It requires a strong push from the top down and a re- ciprocal enthusiastic support and participation from the bottom up. Data analytics is part of a university’s decision fabric that requires stra- tegic planning from an institutional perspective and the allocation of resources that reflect its growing importance in support of the institu- tion’s mission and vision for the future. To be successful in instituting a data- informed decision culture, there are three main conditions that en- able DIDM to be accepted and practiced in the higher education envi- ronment. They are the people, the technology, and the pro cess and culture.
- 29. 12 Technology, Digitization, Big Data People: Leadership and the Analytics Community University leaders have a very impor tant role to play in data- informed decision making. Their commitment, support, and willingness to use data in supporting their decision making are critical factors in ensuring the successful development of a data- informed decision culture. In its Leadership Agenda series, leaders of Achieving the Dream (ATD), a non- profit organ ization advocating for college access and success, urges in- stitutional leaders to set the tone for commitment to data. ATD believes that committed leadership is central to establishing a culture of con- tinuous improvement that is grounded in inquiry and evidence. Presi- dents, department heads, and other institutional leaders should model be hav iors that support a culture of evidence and inquiry throughout the institutions. ATD further believes that institutional leaders should reg- ularly review and explore student outcome data with diverse stakehold- ers in ways that spur thoughtful prob lem solving for student success (Achieving the Dream 2012). Similarly, Long Beach City College Dis- trict (LBCCD) president, Reagan Romali, and her colleagues have made significant pro gress in student success mea sures including degree com- pletions (Toda 2020). More importantly, she and her team have created a culture of exceptional student success, noted as the most improved among all California community college districts in the number of cer- tificates awarded and eighth most improved in the number of degrees awarded (Toda 2020). Institutional leaders can provide support for data analytics develop- ment efforts by relating analytics programs with the university’s strat- egy and vision. In hiring new leaders, institutional officials may find it helpful to ask new leaders about their interest, vision, and experience in using data to support orga nizational growth and per for mance assess- ment. Trustees should hold se nior leaders accountable for delivering accurate, reliable, and comprehensive data for strategy conversations. University leaders can demonstrate their support for DIDM by invest- ing in data talents and analytical capabilities. In 2016, leaders of Lehigh University conducted an organization- wide risk assessment and identified data analytics as a critical gap in
- 30. Data Analytics and Data- Informed Decision Making 13 their orga nizational capabilities. They immediately took action to ap- point the chief information officer and the chief institutional research officer to assem ble a planning team made up of se nior administrative leaders and data stewards to develop a strategic analytics plan. The plan addressed some of the most critical areas of building a DIDM analytics culture, including the data management infrastructure, data governance, data reporting and collaboration, and the sharing of analytical insights. Most impor tant, Lehigh University’s leadership put resources behind these initiatives and enabled the hiring of key personnel and the acqui- sition of new data management and reporting tools. Actions included moving the business intelligence staff to co- locate with the institutional research and analytics staff, setting up a centralized data repository, es- tablishing a Tableau server for generating data reports and data visual- ization, and hiring a data architect and a data governance man ag er.With positive outcomes, leadership support provided the momentum and re- sources that Lehigh University needed to embrace data- informed deci- sion making. Another impor tant base of support for developing a data- informed decision culture is the existence of a critical mass of campus data ana- lytics users and developers who are actively collaborating and sharing their knowledge and skills. Díaz, Rowshankish, and Saleh (2018) be- lieve that analytical talents and users have dif fer ent roles to play and the same individual can play dif fer ent roles depending on the circum- stances. These roles include: • Business leaders: lead analytics transformation across organ ization; • Data engineers: collect, structure, and analyze data; • Data architects: ensure data quality and consistency of pre sent and future data flows; • Workflow integrators: build interactive decision- support tools and implement solutions; • Visualization analysts: visualize data and build reports and dashboards; • Data scientists: develop statistical models and advanced algo- rithms to solve prob lems;
- 31. 14 Technology, Digitization, Big Data • Analytics translators: ensure analytics addresses critical business problems; and • Delivery man ag ers: deliver data and analytics- driven insights and interface with end users. Clearly, as organ izations face the challenge of Big Data, they need analytical talents to help clean the data, or ga nize it, and store it, along with training people to analyze and build models using data. High- performing organ izations tend to support data sharing and encourage collaboration among dif fer ent types of users. A data community is a mutually supportive environment where data users with analytical needs and appropriate security clearance can connect to all available data re- sources across dif fer ent organ ization vectors to detect patterns or con- nections that a single data silo will not help. Mathies (2018) proposes that institutions develop a data sharing mandate, and Arellano (2017) recommends that a data user community be designed as a combination of people across the enterprise, whereas common data and analytical tools are shared. This networked approach helps share information and analytic results across interested groups and those with more skills being seen as a source of trusted analytics for the whole network. This combination of central governance and distributed data access and contribution can help every one get needed information without slow- ing down the business by depending on the central IT team (Arellano 2017). Technology Another critically impor tant enabling condition for DIDM is the avail- ability and access to up- to- date and user- oriented data management and reporting tools, including but not limited to the following core components: • Ability to integrate data from many dif fer ent sources, including but not limited to enterprise resource planning systems (i.e., PeopleSoft, Banner, etc.), third- party software systems, and cloud- based platforms, both internal and external sources;
- 32. Data Analytics and Data- Informed Decision Making 15 • A strong data governance system that helps standardize and systematically document data definitions, data dictionaries, data specifications, and data lineages; • Availability of effective data reporting, data analy sis, and data visualization tools; and • Ability to harness the power of structured, semi- structured, and unstructured data resources through data architecture designs such as a data lake. An enterprise- wide data management and sharing infrastructure typ- ically comes in the form of an enterprise data ware house (EDW). Tra- ditionally, an EDW is installed on- site at the institution in a database server and managed by the IT department. Technological advances in the last several years have allowed organ izations to move EDW opera- tions to the cloud. For Big Data storage, the concept of a data lake or data reservoir may be considered. A data lake is a data management methodology enabled by a massive data repository based on low- cost technologies that improves the capture, refinement, archival, and ex- ploration of raw data within an enterprise. This repository may con- tain unstructured, semi- structured, and structured data where the larg- est part of these data may have unrecognized value for the organ ization (Khine and Wang 2017; Watson 2015). Data lakes are often built by tap- ping into the vast storage space made available by cloud- based com- puting platforms such as Amazon’s or Microsoft’s cloud solutions. The availability of more data and from many more sources not only poses a challenge for storage and access, but also for the documenta- tion and standardization of data ele ments. No matter whether it is in an EDW environment or a cloud- based data lake environment, a data governance structure with strong enforceability is a must. As a collec- tion of practices and pro cesses that help to ensure the formal manage- ment of data assets within an organ ization (Knight 2017), data gover- nance is an orga nizational pro cess that involves other activities such as data stewardship, data quality control, and data security. Together, these activities help an institution gain better control over its data assets, including methods, technologies, and be hav iors around the proper
- 33. 16 Technology, Digitization, Big Data management of data. For more detail, Glasgal and Nestor systematically introduce the concept of data governance and share how the system was implemented at Northeastern University in chapter 6 of this volume. Another technological must for DIDM is the wide adaption of data reporting and visualization tools in sharing data insights with constitu- ent groups, especially with the se nior leadership. Gone are the days when data reports come in with dozens of statistical tables and many pages. With data visualization tools such as Tableau (tableau . com) and PowerBI (powerbi . microsoft . com), data are now shown in dif fer ent graphical formats, fitting the types of data used in the reporting. For example, to report historical trends in college enrollment, instead of using a table with columns and rows, data visualization tools now make the trend displayed in a line or bar graph, with many dif fer ent filters to drill down to dif fer ent colleges and departments and by dif fer ent types of students. When done well, following princi ples of good graphic design, a data visualization page can replace a large number of traditional tables. De- scribed in chapter 4, clear and concise communication is essential. Vi- sualized data reports can deliver the data insights quickly and provide an interactive ele ment that can be more useful than static tables. With newer data- reporting tools, key data reports such as management dash- boards, factbooks, student profiles, and productivity reports can now be made visually attractive and easy to understand. For DIDM, data insights delivered in an easy to understand and easy to access manner are key to ac cep tance and utilization. An example of Lehigh Universi- ty’s enrollment report is given in a visually pleasing and highly intuitive format in figure 1.1. The visualization module enables a user to inter- actively query the data by many layers of data filters: semester, level of students, class of students, race/ethnicity, cohorts, on- campus vs. off- campus, and FTE vs. headcounts. This report replaces many detailed data tables in traditional paper- based or PDF- generated reports.* Another technological advancement in data analytics is the collec- tion and analy sis of social media and human interaction data. This new *Lehigh University’s interactive data visualization tool can be accessed at https://oirsa .lehigh.edu/enrollment.
- 34. Data Analytics and Data- Informed Decision Making 17 approach is best captured in the “connected campus” idea proposed by a number of companies such as Salesforce, Oracle, and Microsoft. Many higher education institutions are data rich and information poor. Insti- tutions collect student data using enterprise resource planning (ERP) systems like Banner or PeopleSoft, but the data are mostly locked behind security layers and not utilized for analytical pro cessing. Officials track high school students who visit institutional web sites, come for campus tours, and submit applications, but in most cases these data are not con- nected to predict and support their future success once they arrive on campus. Rec ords are kept for students who participated in vari ous cam- pus activities, but the data are scattered and not utilized to personalize and enhance students’ learning experience. Academic advisors meet with students regularly but are not equipped with the right data to individu- alize their interactions. They know that degrees are granted, but advisors have limited knowledge about students’ career success and continued en- gagement with their alma maters. Figure 1.1. Lehigh University Enrollment Report (based on Tableau platform). https://oirsa.lehigh.edu/fte-headcount
- 35. 18 Technology, Digitization, Big Data While these data issues may not have been major barriers to student success in the past, institution officials’ ability to improve retention, grad- uation, and lifelong engagement of students depends on improving their “connectedness.” The connected campus idea is based on the customer relations management (CRM) platform (e.g., Salesforce . com), which acts as a communication tool for dif fer ent campus departments to track their interactions with dif fer ent stakeholder groups. A CRM stores data from all sources and organizes it in a way that facilitates personalized commu- nications. For example, an academic advisor armed with a CRM will be able to interact with the student more effectively if he or she can access the student’s academic rec ords, student life, and career development op- portunity data in one place. In chapter 9, O’Bryan explains how college officials can change their level of engagement with students by connecting the disparate data points to understand the full lifecycle of student en- gagement from the time of initial interest in the institution throughout the students’ interaction with the institution before and after graduation. Pro cess and Culture Leadership support, a community of analytics talents, and a strong tech- nology infrastructure are the strong foundation for developing DIDM. To truly make DIDM a success, institution leaders must change their business pro cesses and intentionally build an analytics culture. This cul- tural transformation starts with the articulation of the basic princi ple of treating data as an institutional asset and not a resource owned or monopolized by a department or unit. In a survey of higher education leaders, Bichsel (2012) found that the data silo is a particularly com- mon prob lem in higher education. For an analytics program to be suc- cessful, orga nizational policies must be changed to encourage the shar- ing, standardization, and federation of data resources, balancing the needs for security with needs for access. For DIDM to take root, the following are key considerations: • Se nior leaders need to show commitment to using data to inform decisions by asking for and utilizing data analytics insights;
- 36. Data Analytics and Data- Informed Decision Making 19 • DIDM requires the breaking down of the orga nizational silos to facilitate data sharing and collaboration—no individual unit or department “owns” the data, but rather it is part of the univer- sity’s data resources and needs to be shared based on appropriate security and data governance rules; • IT, institutional research (IR), and operational management should work in close collaboration to explore data and analyze data findings to discover actionable insights; orga nizational leaders must be willing to take the actionable insights to pi lot test new orga nizational change or operational improvement ideas; and • Given the large number of challenges facing higher education institutions, DIDM efforts will add greater value if such efforts can focus on institutional priorities (such as student success). Data silos are often barriers to greater levels of transparency in per for mance assessment and institutional planning. Gagliardi and Turk (2017) point out that the democ ratization of data analytics might reveal some incon ve nient truths about the per for mance of colleges and universities. However, a greater level of data transparency is needed as the higher education sector becomes more competitive and stakehold- ers demand greater accountability. Instead of letting orga nizational silos become barriers to needed changes, college and university leaders should empower change by providing critical operational and per for mance data to key stakeholders so that they can use the shared data resources to make informed decisions. For example, at a private college in the Northeast United States, a collegewide interactive dashboard proj ect got stuck in the implementation phase when the deans and department chairs demanded that their data be kept from other deans and chairs. To meet the needs of the deans and chairs, the complexity of the data classifica- tion schema and access privilege rules increased almost exponentially, making the data programmers’ job a nightmare. Even when the pro- grammers were able to create data visualizations for the reports with multiple layers of administrative access rules, the resulting data reports lost all the connectivity and relative comparisons that a visualization tool
- 37. 20 Technology, Digitization, Big Data is designed to deliver. To truly embrace DIDM, college leaders must break down the data silos and show some courage in enabling data transparency. Enabling data analytics to be embedded in the institution’s culture and be successful will also require plans for training and professional development. Training may be needed for the technical aspects of data storage and maintenance, for analysts who must be deeply knowledge- able of data definitions, techniques for manipulation of the data, and considerations of ethical and responsible uses of data. User groups may be one way to ensure that multiple users across a campus are consis- tent in their understanding of the data and how it applies within their specific contexts. Indeed, training and professional needs are an impor tant part of the institution’s long- term data analytics program, and more on this topic is discussed in chapter 12. Another impor tant aspect of cultural transformation in data analyt- ics is the willingness to give data insights a chance to inform decision making. Leaders must have both the patience and the willingness to let data provide clues, to take some risk, and allow program experimenta- tion. To have an innovation mindset is critically impor tant because Big Data, artificial intelligence (AI), and machine learning (ML) will likely create disruptive changes. For example, one college’s admissions office staff produced a well- designed and detailed glossy brochure to attract more applicants to help achieve its goal of expanding its enrollment for five consecutive years. Admissions officials sought to send the brochure to every applicant who visited their web site and requested additional information. Given the high cost of printing, the vice president for ad- missions de cided to divide the prospects into two groups, with Group A prospects receiving the glossy paper brochure and prospects in Group B receiving a PDF version of the brochure through email with enhanced web- based contents. With the goal of finding out if an electronic bro- chure is equally effective in encouraging application, this experiment came with risk; if the electronic brochure was not well received, the col- lege would have missed its enrollment target. College officials proceeded with the experiment and affirmed that it was a risk worth taking, because they believed that Generation Z students (the primary demographic group
- 38. Data Analytics and Data- Informed Decision Making 21 who are interested in this college) are more receptive to electronic mate- rials. More impor tant, they wanted to use data and results from this quasi- experiment to inform future admissions strategies. Another key aspect of building a data- informed decision environment is the collaboration between the information technology (IT) and the analytics communities. IT is a critical partner that contributes to the strong and dynamic analytical environment of the campus. When asked, “What is your data strategy?” DalleMule and Davenport (2017) argued that a data strategy framework should distinguish between data defense and data offense— each with dif fer ent objectives, activities, and archi- tecture. A defensive data strategy focuses on ensuring data integrity, data security, data access, and data documentation. An offensive data strat- egy centers on generating insights from data to support business pro cess, generate business value, and achieve organ ization objectives. In other words, defense is what IT is good at providing, and offense is what business users and analysts are good at developing. Defense and offense need to work well together to become effective in implementing orga nizational data strategies. All higher education institutions need both offensive and defensive data strategies to be successful in DIDM. The Imperatives for DIDM in Higher Education The Expectation Imperatives of DIDM Many individuals hold high expectations for higher education. Stake- holders such as students and parents expect costs to be controlled, time to degree to be reasonably short, graduation rates to be high, and for employment to be secured soon after graduation. Business leaders ex- pect universities to equip students with employable skills who can con- tribute to prob lem solutions. Government leaders expect universities to operate efficiently and contribute to regional and local economic devel- opment. With these expectations, universities are under scrutiny to prove their value. Many aspects of the university’s operations will need to be supported by strong analytics programs. These include the six items described below:
- 39. 22 Technology, Digitization, Big Data Student Success and Outcomes For all higher education institutions (HEIs), student success and out- comes should be the most impor tant mission. The success of Georgia State University in improving student success using analytical insights (see chapter 8 of this book) is a great example of how DIDM can add value and truly make a great difference. Student success should be a core ele ment of university strategy at the most se nior level of the organ ization. Marketing and communications should highlight student success as a central piece of the institution’s strategic mission. A sustainable plan should include data models and results showing return on investment at an institutional level. As the pro cess scales, retention improvement will help improve revenue stream and improve instructional quality. Leader- ship should consistently communicate a vision of student success— this can, in turn, effectively align resources to support defined goals. New Academic Program and Curriculum Innovation Analytical tools such as learning analytics, customer relations manage- ment, machine learning, and artificial intelligence will create opportuni- ties for new designs of academic programs and through mass customiza- tion. New developments such as stackable credentials, learning badges, and experiential transcripts are more connected with student learning needs and with demands of the job market. Davenport et al. (2001) have pointed out that, armed with Big Data analytics, more organ izations will be able to better understand customers’ needs and will, subsequently, create new products for those needs. Higher education can and should use Big Data analytics to support program innovations and changes that meet the changing needs of the students and employers. Meeting the Needs of the Community and Industry In discussing a university’s relation with external communities Gavazzi and Gee (2018) use spousal relationships as a meta phor to argue that universities must cultivate relationships to have harmonious and pros- perous interactions with their communities. To address the value prop-
- 40. Data Analytics and Data- Informed Decision Making 23 ositions to their community and industry partners, university officials should work proactively to create and sustain programs that are mutu- ally beneficial. In today’s digital age and global competitions, a univer- sity cannot be an ivory tower isolated from its surroundings. Univer- sity missions and programs are connected to communities and industry in large part as students acquire employable skills and knowledge that meet community and industry needs. DIDM will help by informing uni- versity leaders and faculty members about labor market trends, assess- ing students’ learning experience and leadership capabilities, and mea sur ing the effectiveness of dif fer ent pedagogical approaches. Operational Efficiency and Effectiveness One of the biggest opportunities for the higher education sector in leveraging data analytics for decision making is the ability to improve operational efficiency and effectiveness. Big Data technologies, cloud- based solutions, machine learning, and artificial intelligence will make some of the older technologies and costly solutions obsolete. (See chap- ter 11 of this book for Wayt et al.’s discussion and examples of how analytics support financial and business operations in higher education.) For example, enterprise resource planning systems, including human re- sources, finance, research administration, and student information, will no longer need to be installed and operated on premise and bud geted as an expensive capital expenditure, saving a lot of resources and per- sonnel cost. Instead, universities that migrate to new cloud- based solu- tions will be in a better position to allocate bud get IT spending as operating expenses, which is easier to bud get on an annual basis and minimizes cost surges for major upgrades. Data analytics can also help achieve operational efficiency and effectiveness by bringing data trans- parency and disciplines to per for mance assessment. As resources man- agement and outcome mea sures become more accessible through dash- boards and scorecards, the conversation on how to achieve better results and improve collaboration will lead to newer opportunities for shared ser vices and reduction of redundancy.
- 41. 24 Technology, Digitization, Big Data Strategic Agility and Differentiation More so than in the past, the next 10 to 20 years in higher education will test the ability of leaders to steer their institutions strategically. The challenges facing higher education and the rapid changes in the digital revolution and connectivity may bring disruptive innovations at a speed that is faster than anticipated. Se nior leaders in higher educa- tion must identify the strategic challenges facing their institutions. Ques- tions may include: What strengths or unique capabilities differentiate one institution from another?; what new programs are needed in order to stay competitive?; can one recruit the right number of students based on the desired student profiles given the significant demographic shifts to come?; and can one grow the institution’s revenue base without re- lying heavi ly on tuition increases? University leaders and trustees must grapple with these and many other questions in their decision- making pro cess. Marsh and Thariani provide critical insights into these ques- tions in chapter 5. Data Governance, Security, and Ethical Considerations Another imperative for DIDM is the safeguarding, ethical, and respon- sible use of our data resources. It is impor tant that data be used to generate analytical insights to inform decisions. It is equally impor tant that this is done in a manner that protects the privacy and rights of our students and employees. Chapter 6 addresses impor tant points related to data use and governance, and chapter 4 shares impor tant insights on responsible and secure use of data. Prinsloo and Slade (2015) remind us that the traditional paternalistic HEI culture, along with the more recent enthusiasm for pos si ble enhanced student success through ana- lytics, has influenced attitudes and policies on data collection but has not adequately addressed issues of privacy. Strong data governance and a thorough plan for safe collection and storage of data are critical. Cloud- based solutions and the proliferation of third- party applica- tions will continue to create challenges for data management. Most of the policy and pro cess questions need to be addressed through a data governance body to ensure legal and regulatory compliance and to re-
- 42. Data Analytics and Data- Informed Decision Making 25 duce organ ization risk exposure. Similarly, as more data resources are being used to create predictive models and algorithms that impact stu- dents’ lives and outcomes, greater attention and care need to be taken to ensure that the privacy rights of the study subjects are being safeguarded. In chapter 4, Webber and Morn also address some of the human factors and subjective judgments needed in the use of data. Many decisions re- quire careful calibration of the po liti cal, financial, and social factors. DIDM is a cultural change and not a one- time proj ect. For DIDM to work well, university leaders and the user community need to embrace it as a platform and a culture, not a proj ect that needs to be completed. DIDM is not just about the data tools or the newer technologies, it is more importantly about the data awareness and analytical insight ac cep tance and utilization mindset. Enabling this change will also require a strategy for personnel training for analytic techniques. EDUCAUSE (2012) recommends that higher education leaders ask the right strate- gic and operational decision questions and seek to use data evidence to answer these questions and find the right solutions: invest in data tal- ents and data insight translators and foster a vibrant data user com- munity on campus; do not let perfection be the enemy of data uses; make the best out of available data information resources; encourage closer collaboration between the IT and the analytics communities; and invest in analytical tools and technologies that will facilitate the inte- grated view of data insights across the campus. Conclusion Advances in technology, including storage for large volumes of data, are challenging the ways in which decisions are made in higher education. Nearly all, if not all, stakeholders desire more data, assuming that it will make better decisions. We believe that, unlike data- driven methods that rely heavi ly on predetermined algorithms, data- informed decision mak- ing will facilitate goal completion and help achieve greater effective- ness for higher education institutions. DIDM involves both top- down commitment and bottom-up support, strategic planning and resources that acknowledge the institution’s mission and vision for the future, and
- 43. 26 Technology, Digitization, Big Data lots of hard work. A strong foundation for DIDM rests on leaders who support and facilitate orga nizational programs and procedures that de- velop and build a community of analytics talents. University leaders have a critical role to play in data- informed decision making; their com- mitment, support, and willingness to use data to support decision mak- ing is among the most critical factors that will ensure the successful im- plementation of a data- informed decision culture. Although the volume and variety of data continue to increase at a faster speed, institutional leaders as well as external stakeholders must consider the practical and ethical uses of data in higher education as they strive to stay ahead of the data tsunami. While vendor products abound, users or potential users should ask hard questions about the “what” can practically be learned from the data as well as the accuracy of the statistical models or algorithms being used. Users must guard against predictive analyses that include subtle biases or produce other unin- tended consequences (Ekowo and Palmer 2017; O’Neil 2016). An in- stitution’s comprehensive data governance plan is incredibly impor tant. Officials may wish to review Mathies’s (2019) proposed Data Bill of Rights, which requires a plan to protect individual data as well as practices that promote data definitions, rules of use, transparency, and shared governance. Many aspects of the university’s operation will benefit from a strong analytics program. Building partnerships with the local community and businesses and ensuring strong data governance and privacy policies are key drivers to the further advancement of data analytics in higher edu- cation that will facilitate student and institutional success.Analytic strat- egies of data will not be minimized, only further emphasized as we move forward. References Achieving the Dream. 2012. “Building Institutional Capacity for Data- Informed Decision Making.” Public Agenda Series #3. Accessed December 31, 2018. https:// www.achievingthedream.org/sites/default/files/resources/ATD_CuttingEdge_No3.pdf. Arellano, P. 2017. “Making Decisions with Data— Developing a Community around Data in Your Business.” IT Pro Portal. Accessed Feb. 21, 2019. https:www .itproportal.com/features/making-decisions-with-data-developing-a-community -around-data-in-your-business/.
- 44. Data Analytics and Data- Informed Decision Making 27 Arnold, K., and M. Pistilli. 2012.“Course Signals at Purdue: Using Learning Analytics to Increase Student Success.” Proceedings of the Second International Learning Analytics and Knowledge Conference (pp. 267–270). ACM Publications. Beckhard, R. 1969. Organ ization Development: Its Nature, Origin, and Prospects. Reading, MA: Addison-Wesley. Bichsel, J. 2012. “Analytics in Higher Education: Benefits, Barriers, Pro gress, and Recommendations.” Louisville, CO: EDUCAUSE Center for Applied Research. http://www.educause.edu/ecar. Brown, M. 2011. “Learning Analytics: The Coming Third Wave.” (EDUCAUSE Learning Initiative Brief). Retrieved from EDUCAUSE library. https://library .educause.edu/resources/2011/4/learning-analytics-the-coming-third-wave. Chin, J. K., M. Hagstroem, A. Libarikian, and K. Rifai. 2017. “Advanced Analytics: Nine Insights from the C- Suite.” McKinsey Analytics (July): 1–8. Chai, S., and W. Shih. 2017. “Why Big Data Isn’t Enough.” MIT Sloan Management Review (Winter): 57–61. Chris Brady, C., M. Forde, and S. Chadwick. 2017. “Why Your Com pany Needs Data Translators.” MIT Sloan Management Review (Winter): 14–16. DalleMule, L., and T. Davenport. 2017. “What’s Your Business Strategy?” Harvard Business Review (May– June). Accessed December 15, 2018. https://hbr.org/2017 /05/whats-your-data-strategy. Davenport, T. H., J. G. Harris, D. W. De Long, and A. L. Jacobson. 2001. “Data to Knowledge to Results: Building Analytics Capability.” California Management Review 43, no. 2: 117–138. Díaz, A., K. Rowshankish, and T. Saleh. 2018. “Why Data Culture Matters.” McKinsey Quarterly (September). Accessed April 2019. https://agribusiness .purdue.edu/quarterly-review/why-data-culture-matters. Drake, B. M., and A. Walz. 2018. “Evolving Business Intelligence and Data Analytics in Higher Education.” In New Directions for Institutional Research, 178, edited by C. Mathies. San Francisco: Wiley Periodicals, Inc. Drachsler, H., and W. Greller. 2016. “Privacy and Analytics— It’s a DELICATE Issue: A Checklist to Establish Trusted Learning Analytics.” Proceedings of the Sixth International Conference on Learning Analytics and Knowledge (LAK ’16), 89–98. http://dx.doi.org/10.1145/2883851.2883893. Dziuban, C., P. Moskal, T. Cavanagh, and A. Watts. 2012. “Analytics That Inform the University: Using Data You Already Have.” Journal of Asynchronous Learning Networks 16, no. 3: 21–38. Ekowo, M., and I. Palmer. 2017. “Predictive Analytics in Higher Education: Five Guiding Practices for Ethical Use.” Educational Policy Paper. Washington, DC: New America. Frick, W. 2014. “An Introduction to Data- Driven Decisions for Man ag ers Who Don’t Like Math.” Harvard Business Review. May 19, 2014. https://hbr.org/2014/05/an -introduction-to-data-driven-decisions-for-managers-who-dont-like-math. Gagliardi, J. S., and J. M. Turk. 2017. The Data- Enabled Executive: Using Analytics for Student Success and Sustainability. Washington, DC: American Council of Education. Gavazzi, S. M., and E. G. Gee. 2018. Land- Grant Universities of the Future. Baltimore, MD: Johns Hopkins University Press.
- 45. 28 Technology, Digitization, Big Data General Accounting Office. 2016. “Data and Analytics Innovation— Emerging Opportunities and Challenges.” September 2016. GAO-16-659SP. Gharib, S. 2018. “Intel CEO Says Data Is the New Oil.” Fortune Magazine. June 7, 2018. http://fortune.com/2018/06/07/intel-ceo-brian-krzanich-data/. Goldring, E., and M. Berends. 2009. Leading with Data Pathways to Improve Your School. Thousand Oaks, CA: Corwin Publishing. Grajek, S., and the EDUCAUSE IT Issues Panel. 2019. “Top 10 IT Issues, 2019: The Student Genome Proj ect.” Washington, DC: EDUCAUSE. Heavin, C., and D. J. Power. 2017. “How Do Data- Driven, Data- Based and Data- Informed Decision Making Differ?” Blog post. Accessed February 11, 2019. http://dssresources.com/faq/index.php?action = artikel&id = 392#. Jaschik, S., and D. Lederman. 2019.“The 2019 Inside Higher Education Survey of College and University Chief Academic Officers— A Study by Gallup and Inside Higher Education.” Inside Higher Education. January 23, 2019. https://www .insidehighered.com/news/survey/2019-inside-higher-ed-survey-chief-academic -officers. Khalil, M., and M. Ebner. 2015. “Learning Analytics: Princi ples and Constraints.” In Proceedings of World Conference on Educational Multimedia, Hypermedia and Telecommunications 2015, pp. 1326–1336. Chesapeake, VA: AACE. Khine, P. P., and Z. S. Wang. 2017. “Data Lake: A New Ideology in Big Data Era.” Paper presented at the Fourth International Conference on Wireless Communica- tion and Sensor Network, Wuhan, China, December 2017. Accessed February 13, 2019. https://www.researchgate.net/publication/321825490_Data_Lake_A_New _Ideology_in_Big_Data_Era. Knapp, M. S., M. A. Copland, and J. A. Winnerton. 2007. “Understanding the Promise and Dynamics of Data- Informed Leadership.” In Yearbook of the National Society for the Study of Education: 74–104. Knight, M. 2017. “What Is Data Governance?” Dataversity. Blog post. December 18, 2017. Accessed Feb. 13, 2019. https://www.dataversity.net/what-is-data-gover nance/. Lane, J. E., and B. A. Finsel. 2014. “Fostering Smarter Colleges and Universities— Data, Big Data, and Analytics. In Building a Smarter University: Big Data, Innovation, and Analytics, 3–26, edited by J. E. Lane. Albany, NY: State Univer- sity of New York Press. Marr, B. 2018. “Forget Data Scientists and Hire a Data Translator Instead?” Forbes Magazine Online. March 12, 2018. https://www.forbes.com/sites/bernardmarr /2018/03/12/forget-data-scientists-and-hire-a-data-translator-instead/#48ade0 b0848a. Martin, D. 2016. “Data- Driven versus Data- Informed— What’s Best for You?” Cleverism. August 12, 2016. https://www.cleverism.com/data-driven-versus-data -informed-whats-best/. Mathies, C. 2018. “Ethical Use of Data.” In IR in the Digital Era: New Directions for Institutional Research 178: 85–97, edited by C. Mathies and C. Ferland. Boston: Wiley. Maycotte, H. O. 2015. “Be Data- Informed, Not Data- Driven, for Now.” Forbes Magazine Online. January 13, 2015. https://www.forbes.com/sites/homaycotte /2015/01/13/data-informed-not-data-driven-for-now/#7ccdebc6f5b7.
- 46. Data Analytics and Data- Informed Decision Making 29 McAfee, A., and E. Brynjolfsson. 2012. “Big Data: The Management Revolution.” Harvard Business Review (October). https://hbr.org/2012/10/big-data-the -management-revolution. O’Neil, C. 2016. Weapons of Math Destruction: How Big Data Increases In equality and Threatens Democracy. New York: Broadway Books. Peterson, R. J. 2012. “Policy Dimensions of Analytics in Higher Education.” EDUCAUSE Review (July– August): 44–47. Pfeffer, J. 1998. “The Human Equation: Building Profits by Putting People First.” Cambridge, MA: Harvard Business School Press. Picciano, A. G. 2012. “The Evolution of Big Data and Learning Analytics in American Higher Education.” Journal of Asynchronous Learning Networks 16, no. 3: 9–20. Prinsloo, P., and S. Slade. 2015. “Student Privacy Self- Management: Implications for Learning Analytics.” In Proceedings of the LAK 2015 International Conference on Learning Analytics and Knowledge, 83–92. New York, NY: Association of Computer Machinery. Provost, F., and T. Fawcett. 2013. “Data Science and Its Relationship to Big Data and Data- Driven Decision Making.” Big Data 1, no. 1: 51–59. https://doi.org/10.1089 /big.2013.1508. Reinitz, B. T. 2015. “Building Institutional Analytics Maturity” (Summit report). Louisville, CO: EDUCAUSE Center for Analy sis and Research. Salesforce. 2019. “Software Com pany for Customer Relations Products and Tools.” Salesforce. Accessed February 2, 2019. https://www.salesforce.com/crm/. Smolan, R., and J. Erwitt. 2012. The Human Face of Big Data. Sausalito, CA: Against All Odds Productions. Swing, R., and L. Ross. 2016. “A New Vision for Institutional Research.” Change 48, no. 2: 6–13. https://doi.org/10.1080/00091383.2016.1163132. Toda, S. (2020). Long Beach city college makes phenomenal pro gress in students earning degrees and certificates. Press Release from Digital Journal, February 4, 2020. Accessed February 12, 2020. http://www.digitaljournal.com/pr/4578138 Viberg, O., M. Hatakka, O. Balter, and A. Mavroudi. 2018. “The Current Landscape of Learning Analytics in Higher Education.” Computers in Human Behavior 89: 98–110. Waston, H. J. 2015. “Data Lakes, Data Labs, and Sandboxes.” Business Intelligence Journal 20:4–7. White, A. 2018.“Do You Really Want to Be Data- Driven?” CIO Magazine. Novem- ber 21, 2018. https://www.cio.com.au/article/649883/do-really-want-data-driven/. Winkler, M. K., and S. D. Fyffe. 2016. “Strategies for Cultivating an Orga nizational Learning Culture.” Urban Institute White Paper. December 2016. https://www .urban.org/sites/default/files/publication/86191/strategies_for_cultivating_an _organizational_learning_culture_3.pdf. Zheng, H. Y. 2015. “Business Intelligence as a Data- Based Decision Support System and Its Roles in Support of Institutional Research and Planning.” In Institutional Research and Planning in Higher Education— Global Contexts and Themes, 159–173, edited by K. Webber and A. Calderon. New York: Routledge. Zwitter, A. 2014. “Big Data Ethics.” Big Data & Society 16: 1–6. https://doi.org/10 .1177/2053951714559253.
- 47. 30 Gosh, you’ve really got some nice toys here. —Roy Batty, Blade Runner (1982), set in 2019 In the summer of 1956, luminaries of mathe matics and information sciences gathered at Dartmouth College for two months to hold in- depth discussions about what the group or ga nizer John McCarthy termed artificial intelligence. McCarthy’s proposal to the Rocke fel ler Foundation to support the summer meeting provocatively and pre- sciently asserted: “If a machine can do a job, then an automatic calcula- tor can be programmed to simulate the machine. The speeds and mem- ory capacities of pre sent computers may be insufficient to simulate many of the higher functions of the human brain, but the major ob- stacle is not lack of machine capacity, but our inability to write pro- grams taking full advantage of what we have” (McCarthy et al. 1955). Participants in this summer meeting included, among others, Arthur Samuel, who would later in the de cade coin the term machine learning as he developed a computer program that could win a checkers game against a human being; Ray Solomonoff, who developed algorithmic information theory that machine learning is probabilistic and can be trained on existing data to solve new prob lems; Marvin Minsky, who would go on to co- found the AI Lab at MIT; and Claude Shannon at [ 2 ] Big Data and the Transformation of Decision Making in Higher Education Braden J. Hosch
- 48. The Transformation of Decision Making 31 Bell Labs, who developed information theory and would later in life quip that in computer- humans chess games, he was “rooting for the ma- chines” (Shannon 1987). The meeting was significant not just because of the brilliance of its attendees but because of the prob lems addressed, which included discussion on automatic computers, how computers can be programmed to use a language, neuron nets, theory of the size of a calculation, self- improvement of data programs, abstractions, and ran- domness and creativity. These prob lems represented the central challenges of achieving artificial intelligence envisioned by Alan Turing (1950) sev- eral years earlier that a machine could mimic human be hav ior. Solutions, however, remained elusive until improvements were made to computing power, data storage and management systems, and net- working. The resulting developments are changing and will continue to change how organ izations, including universities, operate and deliber- ate. This chapter provides historical context for how the transforma- tion of computing from rec ord keeping and administrative pro cessing into what Agrawal, Gans, and Goldfarb (2018) call “prediction ma- chines” affects how decisions are made and how Big Data represent a transformational means for colleges and universities to improve if not reimagine their operations. Big Data, in this re spect, is a broad term that includes huge amounts of structured data (such as all the clicks of all students in all online learning materials at a university), but also unstructured data like social media feeds with text, images, and video files, as well as a set of non- hypothesis- driven analytical techniques ap- plied to existing (smaller) data sets. This chapter asserts that resulting developments in machine learning, artificial intelligence, and the Inter- net of Things provocatively point toward a future for the higher educa- tion sector in which decisions made by students, faculty, and adminis- trators are approached much differently from earlier periods. The Evolution of Computing Power and University Decision Making In the 1950s, colleges and universities were organ izations of students, faculty, and staff, concentrated on a geo graph i cally defined campus, and
- 49. 32 Technology, Digitization, Big Data who labored to produce voluminous textual material— articles, books, term papers, tests, and memos. World- class science occurred in labora- tories, but results were recorded in quadrille notebooks and written up as lab reports, later typed up, submitted to and published in journals that would later be housed on the bookshelves of libraries. The walls of registrar’s offices were obscured by beautiful wooden filing cabinets with reams of student files printed on paper (often handwritten) and stored in actual folders. Decisions about whom to admit or not, whom to hire or let go, what programs to start or retire were made largely on the basis of professional expertise and the judgment of experts who had spent entire careers at an institution. As Gladwell (2005) demonstrates, there is real value in the judgment of experts in their fields of expertise, but these judgments are also necessarily bounded by the knowledge of those making them. Administrative computing became a real ity at research universities following its deployment in academic computing in the late 1950s and 1960s, featuring large mainframe computers built by IBM and later Bur- roughs and Cray running with vacuum tubes adjacent to large cooling facilities. Initial pro cessing power in the mid-1950s was mea sured in hundreds of instructions per second, increasing to millions of operations per second by the late 1960s (IBM 2003), several orders of magnitude slower than the personal mobile devices of the late 2010s, which rec ord speeds of billions of operations per second (Simonite 2018). Data analy sis became easier with the release of statistical software applica- tions still in use today, such as the Statistical Package for the Social Sci- ences (SPSS) released in 1968 and the Statistical Analy sis System (SAS) released in 1971. As faculty circulated through administrative roles, in- cluding the relatively new function of institutional research, these ap- plications became widespread tools of choice among institutional re- searchers to prepare descriptive statistics informing institutional lead- ership about the past. This knowledge was invaluable to institutional decision making, and campus planning estimates made use of cohort attrition models for enrollment planning and segmented yield rates for admissions, but forecasting still relied heavi ly on professional expertise informed by population- level statistics.
- 50. The Transformation of Decision Making 33 Even though computing power was still relatively limited, the prom- ise of artificial intelligence to transform education was under active ex- ploration, as evidenced by Ellis Page’s initiative (Page, Fisher, and Fisher 1968) to grade composition papers using the computing power of the day. Proj ect Essay Grade, funded by the US Department of Education (Page and Paulus 1968), investigated the feasibility of automatically analyzing and evaluating student writing using a FORTRAN program for natu ral language pro cessing after student papers were keyed into mainframes. Proj ect Essay Grade demonstrated that computer programs were about as good as human raters at evaluating student writing, al- though the methods remained too costly for widespread adoption (Page, Fisher, and Fisher 1968). Page later revived the proj ect in the 1990s, and with the exponential increase in computing power, the widespread use of computer terminals in testing, and the motivation of testing companies to cut costs, the basic infrastructure of Page’s proj ect became ubiquitous in the 2000s. Data storage and pro cessing also evolved markedly during the 1950s and 1960s, and the increased capacity to store data had implications for decision- making pro cesses. Data and programs were created and stored on punch cards— technology from the nineteenth century to au- tomate textile production. Use of magnetic tape to store data was in- troduced by IBM in 1951 and offered great advantages for increasing speed and volume but still carried the limitations of sequential storage. In the mid-1950s and with marked advancements in the 1960s, hard disks allowed for random access to the blocks in which data were stored, providing additional advances in storage capacity and retrieval speed. Importantly, the technology allowed development and commercializa- tion of the floppy disk in the late 1960s and early 1970s, which allowed for the transport of data between microcomputers and mainframes. Di- rect access storage of data versus sequential storage on tape or a box of punch cards also allowed for development of data management systems, with the introduction of navigational databases in the 1960s and the relational database in the next de cade by Edgar Codd (1970). These events were followed by the development of structured query language (SQL) later in the de cade (Chamberlin and Boyce 1974), later to be
- 51. 34 Technology, Digitization, Big Data commercialized by Oracle for release in 1979 and still in widespread use forty years later. These intensive mainframe computing resources and data tools, how- ever, were generally reserved for large research universities, not smaller colleges; it was not until the 1980s, with the proliferation of the micro- computer, or personal computer (PC), into faculty and administrative offices, that computing power became inexpensive enough to become widespread for management of colleges and universities. Gilbert and Green (1986) describe this era as the computing revolution, noting that almost half a million microcomputers were operating on campuses by the middle of the 1980s and over half of entering freshmen reported having occasionally or frequently written a computer program. Importantly, per- sonal computers effectively pushed the ability both to generate and access data to every member of the university community, although the poten- tial of this breakthrough was realized only over the succeeding de cades. Gilbert and Green (1986) offered to campus leaders an overview of the challenges and opportunities of technology adoption as well as a taxon- omy for making decisions about technology. However, their focus, and indeed the focus of administrative IT of the period, rested on how col- leges and universities could and should manage information technology while remaining silent about how the computer revolution had the poten- tial to improve management of colleges and universities. The Advent of Enterprise Systems Potential for more widespread application of computing power to man- age the higher education enterprise advanced significantly in the 1980s and 1990s with the migration from locally developed administrative computing systems to broader adoption commercial enterprise resource planning (ERP) systems like Banner and PeopleSoft. University ERP sys- tems brought together many of the basic business operations of univer- sities into an integrated platform, so that registration and student rec ords, billing, bud geting, and human resources management became entirely digitized pro cesses, with data stored in common locations. These systems still notably omitted many mission- level functions of colleges
- 52. The Transformation of Decision Making 35 and universities such as management of learning outcomes, teaching ef- fectiveness, use of student ser vices, and research activity and outcomes. In fact, the absence of these features within major higher education ERP systems has been the hob goblin of efforts to mea sure and improve in- stitutional effectiveness over the past two de cades. Where ERP sys- tems fell short for specific higher education functions, other vendors stepped into the breach with customer relations management (CRM) systems for admissions, learning management systems (LMS) for teach- ing and learning, assessment management systems for educational out- comes, and donor management systems for alumni affairs and advance- ment functions. (For more information on CRM systems, see chapter 9; for more information on LMS systems as part of learning analytics, see chapter 10.) Nevertheless, the ERP systems were transformative for decision- making pro cesses within the institution. Significant and insignificant transactional details about students and employees migrated from pa- per rec ords or siloed spreadsheets to centralized repositories of digital rec ords, every added and dropped course, every salary increase or ex- tra ser vice payment, every purchase and payment was assigned an ef- fective date and stored as a row in a relational database for later re- trieval. From these systems, IR offices, finance and bud get offices, planning offices, and others extracted material for reporting, analy sis, and forecasting. Decision making became reliant upon a culture of re- porting that offered answers to questions in close to real time: How many applicants do we have now compared to the same time as last year? Is our spending for the month in each unit above or below what was bud geted? How many grant applications and for how much money do we have this year compared to the same time last year? Armed with this level of information, university leaders have been better able to ad- just tactics and strategy to respond to current situations. Pro cesses to access, analyze, and communicate this information to leadership are nei- ther automatic nor systemically available, and they required human talent to extract data and transform it into information. Decision mak- ers at institutions with the resources to invest in personnel devoted to analy sis received better intelligence than those who did not.
- 53. 36 Technology, Digitization, Big Data The data ware house also came of age in the 1990s as a response to the proliferation of data from transactional systems, which often yielded conflicting reports to se nior officials because of issues of timing, differ- ing and siloed analyst expertise, and imprecision in how questions were formulated. Bill Inmon (1992) offered the vision that a data ware house could provide an organ ization with a “single version of the truth,” and the star schema for warehousing introduced by Kimball and Merz (2000) became a standard still widely in use. From these systems, busi- ness intelligence (BI) units emerged on many campuses to provide data for decision support. BI units have generally been housed in university IT departments and typically provide a data and reporting infrastruc- ture for client units across campus (Drake and Walz 2018). In some col- leges and universities, this function is fulfilled by institutional research, in others institutional research is a client of the BI unit, and in some instances IR and BI units compete in providing information to other constituencies. In recent years, one approach has been to combine IR and BI units, and as Childers (2016) observes in an orga nizational and anthropological case study of such a merger at the University of Arizona, opportunities for synergy can be counterbalanced by cultural and disci- plinary differences among personnel and even unit missions. Setting the Stage for AI Three subsequent advances led to the explosion of data in the last two de cades that have set the stage for aggressive and increasingly preva- lent use of machine learning and AI: near universal internet coverage, ubiquitous handheld devices, and the use of these devices for social me- dia, internet access, and mobile applications. In the 1990s, transporta- tion of data was accomplished through hard- wire connections on- campus, and at times via floppy disk and slower dial-up connections across campuses. In the following de cade, extensive deployment of high- speed optic cable and high- speed internet access made sharing of larger data files con ve nient and cost effective, especially as creation of appli- cation programming interfaces (APIs) became standard practice among system developers. Satellite networks and mobile towers also contrib-
- 54. The Transformation of Decision Making 37 uted to increased connectivity to support the second key advance: the advent of the smartphone. Since the launch of the iPhone in 2007, which extended the email functionality of the BlackBerry to full internet and web- browsing access, an estimated 5.1 billion unique mobile users were active in 2018, with over 4 billion of them accessing the internet (Kemp 2018). The astonishing magnitude of this number of users becomes dwarfed when considering the amount of data each user generates as he or she browses web sites; accepts tracking cookies; allows data sharing among organ izations; and provides data through “private” forms, trans- actions, and public posts. Effectively, every interaction even down to the click and keystroke is digitized and becomes data that AI needs to con- struct models and make predictions. It is this revolutionary social and transactional feature of the internet, enabled by Google, Facebook, and Twitter, that opened the world of Big Data to global corporate giants as an ave nue to generate profits. And on a smaller scale, the university, through its administrative systems, LMS, and web site, collects data on students, faculty, staff, and visitors— data that are now available to iden- tify patterns and to predict future outcomes. It is impor tant to recognize that data collection is more prevalent than user- system interaction. Large stores of passive data are also being collected, if not yet substantively used, including digital video files from hundreds if not thousands of video security cameras, location and time tracking from the nodes of the wireless network, and repositories of license plates photographed with time stamps of vehicles that enter and exit parking facilities. This amount of data exceeds the capacity and design of the tradition- ally structured data ware house, and now, approaching 2020, college and university officials find themselves at the cusp of moving to more flexible data environments.Web 2.0 companies like Google, Facebook, and Ama- zon shifted away from data ware houses with the snowflake, or star, schema to data environments allowing distributed storage of disparate data types. These platforms, commercially available as products like Ha- doop, SANA, and Amazon Web Ser vices (AWS), offer a No- SQL environ- ment in a nonrelational database, allowing for storage of unstructured data (e.g., Twitter feeds, video files, course assignments from the LMS) alongside structured data. The organ ization reflects an environment
- 55. Another Random Scribd Document with Unrelated Content
- 56. Half an hour more went by; and then was heard the sound of many feet passing along through some chamber near. At the end of above five minutes the door opened, and Monsieur de Tronson led in an elderly lady habited as if for a journey. "Madame de Langdale," said the secretary of the cabinet, addressing Lucette, "Madame de Lagny, with whom you passed last night, will have the pleasure of accompanying you and Monsieur de Langdale on your journey. The carriage has been ready for an hour; but, the council having sat later than usual, I could not leave my post. Monsieur will do me the honor of accompanying me to his chamber below, where I will put him in possession of his money and his safe- conduct, together with his baggage, while you prepare for travelling, which, as it is, must, I fear, be protracted into the night." Edward followed him down several flights of steps, conversing with him, as he went, upon the arrangements for their journey, telling him that he feared from his servant's information they would be obliged to proceed beyond Niort to St. Martin des Rivières, and that, consequently, at least two days more than he had calculated upon must pass ere he could fulfil the promise he had given to return. But De Tronson seemed thoughtful and absent; for, in truth, he had just come from a painful scene;[3] and, although he heard, and answered all his young companion said, it was by an effort, and evidently without interest. All the arrangements were soon made, however. Edward's property was restored to him; the tradesmen he and Lucette had employed were paid; and then the secretary led him to the little court, where stood one of the large clumsy carriages of the day with four tall horses. A stout man on horseback was also there, holding by the rein the horse which Jacques Beaupré had ridden to Nantes, and, as no beast had been provided for Pierrot, he mounted beside the coachman. Lucette and her companion were already in the vehicle, and, with a kind adieu from M. de Tronson, Edward took his place beside them, and the vehicle rolled on.
- 58. CHAPTER XXI. It was a beautiful evening in July, the sky flecked with light clouds just beginning to look a little rosy with a consciousness that Phœbus was going to bed. They cannot get over that modest habit; for, although they have seen the god strip himself of his garmenture of rays and retire to rest every day for—on a very moderate calculation —six or seven thousand years, they will blush now and then when they see him entering his pavilion of repose and ready to throw off his mantle. There is much pudency about clouds. All other things get brazen and hardened by custom, but clouds blush still. It was a beautiful evening in July when the carriage which contained Lucette, Edward, and Madame de Lagny arrived in sight of the chateau of St. Martin des Rivières; but, when they did come in sight, how to get at it became a question of some difficulty. There, on a little mound, stood the building,—not large, but apparently very massive and well fortified,—within a hundred yards of the confluence of two deep and rapid rivers, the passage of each commanded by the guns on the ramparts and on the keep. No bridge, no boat, was to be seen, and for some time the party of visitors made various signals to the dwellers in the chateau; but it was all in vain, and at length Edward Langdale resolved to mount the good strong horse of Jacques Beaupré and swim the nearest stream. Educated in a city, it was not without terror and a sweet, low remonstrance that Lucette saw her young husband undertake and perform a feat she had never seen attempted before; but Edward, though borne with his horse a good way down the stream by the force of the water, reached the other side in safety, and his companions could see him ride to the draw-bridge and enter the castle.
- 59. During some twenty minutes nothing further could be descried; and then, at a point where one of the outworks came down to the river, what I think was called in those days a water-gate was opened, and a boat shot out with two strong rowers. Edward Langdale himself did not appear; but one of the boatmen walked up to the carriage and informed the ladies that his lord, the Duc de Rohan, would be happy to receive them in the chateau, but that the carriage and the men must remain on that side of the river, as the boat could only contain four persons and none other could be had. "Ah, that is the reason Monsieur de Langdale did not return for us," said Madame de Lagny, with whom Edward had become a great favorite. "I was sure he had too much politeness to send servants for his lady if he could come himself." A few minutes passed in placing Lucette's little wardrobe in the boat, and then, with a heart somewhat faint and sad, she followed Madame de Lagny to the water-side, remembering but too acutely that on the opposite bank she was to be received by persons who, however near akin, were but strangers to her, and there, too, very soon to part from him whom she was not now ashamed to own to herself she loved better than any one on earth. The boat shot off from the shore, and though carried so far down by the force of the current that the water-gate could not be reached, yet after some hard pulling the shore was gained, and the two ladies turned toward the drawbridge over which they had seen Edward Langdale pass. Madame de Lagny looked toward the great gate, but the young husband did not appear. In his place, however, was seen a stout middle-aged man, with hair somewhat silvered, and his breast covered by a plain corslet of steel. There were two or three other persons a step farther under the arch; and Madame de Lagny whispered, "That must be the duke himself. But where can Monsieur Edward be?"
- 60. Lucette's heart was asking her the same question; but by this time the Duc de Rohan was advancing to meet her and her companion, and in a moment more he was near enough to take Madame de Lagny's hand and raise it courteously to his lips. "You have come to a rude place, madame," he said, "and among somewhat rude men; but we must do what we can to make your stay tolerable." "Oh, my lord duke," replied the lady, with a courtly inclination of the head, "I must away as soon as possible. I am expected back at the court directly. But where is Monsieur de Langdale? I do not see him." "He is in the chateau, madame," replied the duke; "but he has been telling me so strange a tale that I have judged it best, before he and this—["girl," he was in the act of saying; but he checked himself, and substituted the words "young lady"]—before he and this young lady meet again, to have from her lips and from yours what are the facts of the case. Pray, let us go in." "The facts of the case are very simple, my lord," replied the old lady, with some stiffness. "Monsieur de Langdale is the husband of this young lady, formerly Mademoiselle de Mirepoix, whom you do not seem to recognise, my lord duke, though she is your near of kin. He married her in the presence of the cardinal and the whole court." "More impudent varlet he!" exclaimed the duke, angrily. "And you, mademoiselle,—what have you to say to all this fine affair? Why, you are a mere child! This marriage can never stand!—without any one's consent! It is a folly!" "Not at all, duke," said Madame de Lagny. "Pray, recollect, sir, that Madame de Rambouillet was married at twelve,—I myself at sixteen. Madame is nearly fifteen, she tells me; and, as to the marriage not standing, you will find yourself much mistaken. The man who made it is not one to leave any thing he undertakes incomplete, as you will discover. They are as firmly married as any couple in the land, and that with the full authority of the king, which in this realm of France
- 61. supersedes the necessity for any other consent whatever. She is a ward of the crown, sir; and her father having died in rebellion is no bar to the rights of the monarch." "Madame, I beseech you, use softer words," said the duke, in a calmer tone. "My good cousin De Mirepoix died in defence of his religion, without one thought of rebellion, and really in the service of his Majesty, whose plighted word had been violated not by himself, but by bad ministers who usurped his name. Make room, gentlemen. This way, madame. We shall find in this hall a more private place for our conference." So saying, he led the way into the large room in the lower story of the keep, and there begged Madame de Lagny to be seated. Lucette he took by the arm and gazed into her face for a moment, saying,— "Yes; she is very like. Here, take this stool, child: we have no fauteuils here. Now, answer my question. What had you to do with this marriage? Did it take place at his request or yours?" Lucette's heart had at first sunk with alarm and disappointment at the harsh reception she had received, having little idea what a chattel—what a mere piece of goods—a rich orphan relation was looked upon amongst most of the noble families of France. But the very harshness which had terrified her at first at length roused her spirit; and, though she colored highly, she replied, in a firm tone, "At neither his request nor mine, my lord." "Ah! good!" cried the duke. "Then neither of you consented? The marriage of course——" "We did both consent," said Lucette, interposing. "Did he not tell you the circumstances? Did he not give you the cardinal's message?" "He told me a good deal, and he said something about the Eminence; but, by my faith, I was so heated by the tale that I did not much attend to the particulars. Let me hear your story, mademoiselle. What did the cardinal say?"
- 62. "My lord, we had been stopped near Mauzé by some of the royal officers, and sent on under guard toward Nantes——" "Oh, I know all about that," interrupted the duke. "What have you been doing since? I trust, not masquerading about Nantes dressed up as a page; though, by my faith, ladies are now getting so fond of men's clothes that they will soon leave us none to wear ourselves. Why, there was my good cousin De Chevreuse, with her young daughter, rode across the country, both in cavaliers' habits, and, finding no other gîte, stayed all night with the good simple curé of the parish, who never found out they were women till they were gone. Well, where have you been, and what have you been doing, since that affair at Mauzé?" "The Abbey de Moreilles was burned by lightning, my lord," replied Lucette, whose cheek had not lost any part of its red from De Rohan's language. "We escaped into the Marais, where I was taken ill of the fever common there. As soon as I could travel, we went direct to Nantes, intending to come round at once and seek for Monsieur de Soubise. In consequence of his having sent a man with some of my husband's baggage to that city, we were discovered and arrested." "Your husband, little child?" exclaimed the duke. "But go on; go on. What happened next?" "I was separated from Edward, who had treated me with the kindness of a brother," said Lucette. "Ay, I dare say," again interrupted De Rohan;—"with something more than the kindness of a brother." "For shame, Monsieur le Duc!" said Madame de Lagny, sharply. "You said very truly just now that we had come to a rude place and amongst rude men. If the cardinal had known what sort of reception this poor lady would meet with, I am sure he would have followed the course Monsieur de Tronson hinted at and given her up to
- 63. Madame de Chevreuse. There at least she would have been treated with respect and kindness." At the mere name of Madame de Chevreuse the duke's countenance changed. Without knowing it, good old Madame de Lagny had touched a chord which was sure to vibrate in the heart of any of the Rohan Rohans as soon as one of the Rohan Montbazons was mentioned; and after a moment's pause the prince answered, with a very much less excited air, "His Eminence acted courteously and well in not giving up my fair young cousin to a lady who has no right to her guardianship, who was her father's enemy, whose conduct is not fit for the eyes of a young girl even to witness. But tell me, mademoiselle, what was the message his Eminence sent to my brother to account for his conduct in bestowing—in attempting to bestow—your hand upon an unknown English lad, who may be of good family or may not, but who is no match for any one of the name of Rohan?" "He said, sir," answered Lucette, "that we were to tell you or the Prince de Soubise, whichever we might find, that, under the peculiar circumstances of the case,—by which, I presume, he meant our having travelled so long together,—the cardinal prime minister had judged it imperatively necessary we should be married, and had himself seen the ceremony performed; that for two years Edward should leave me with you, but that at the end of that time he should claim me and take me, and that all his Eminence's power should be exerted to give me to him. He added, in a lower tone, 'They will find me more difficult to frustrate than Madame de Chevreuse.'" "That is true, as I live!" said the duke. "But yet this is hard. Why, girl, it will drive my brother Soubise quite mad,—if he be not mad already, as I sometimes think he is." "His madness will not serve him much against the cardinal," said Madame de Lagny, dryly. "But, my lord, we must bring this discussion to an end, for it is growing dark, and I and Monsieur de Langdale must be treading our way back to Nantes. He is but, as it
- 64. were, a prisoner upon parole; and I promised my cousin De Tronson I would make no delay." "Madame, in all the agitation and annoyance this affair has cost me," said Rohan, "I have somewhat, I am afraid, forgotten courtesy. I ordered refreshments for you, indeed, as soon as I heard of your coming; but I did not remember to ask you to partake of them. They will be here in a moment." "We can hardly stay," said the old lady. "But I beg, sir, you would let Monsieur Edouard be called, both to accompany me and to take leave of his wife." The duke bit his lips; but after a moment's thought he answered, "Pray, madame, take some refreshment. As to this lad, he may come and wish her good-bye; but no private interview, if you please!" The old marquise was a good deal offended at all that had passed, and it was not without satisfaction she replied, "Oh, I dare say they have said all to each other they want to say, Monsieur le Duc. They have had private interviews enough since their marriage to make all their arrangements. Is it not so, dear Lucette?" But Lucette was weeping, and De Rohan, with a cloudy brow, quitted the room. In a few moments some refreshments were brought in and placed upon the table, and the duke appeared, accompanied by Edward Langdale. The youth's look was serious, and even angry, but that of De Rohan a good deal more calm. "Sit down, monsieur, and take some food," said the latter as they entered; but Edward answered at once, "I neither eat nor drink in your house, sir. I did you and your family what service I could, honestly and faithfully; and—because, under force I could not resist, and to save myself and your fair cousin from a fate which you would not have wished to fall upon her nor I wish to encounter for myself, I yielded to a measure which God and she know I never proposed when it was fully in our power—you treat me with indignity. You much mistake English gentlemen, sir, if
- 65. you suppose that such conduct can be forgotten in a few short minutes." "By the Lord!" said De Rohan, with a laugh, "it is well you did not meet with Soubise; for you might have had his dagger in you for half what you have said." "Or mine in him, if he had insulted me further," answered Edward, walking toward Lucette and taking her hand. "A pretty bold gallant," said the duke, with a smile. "Madame de Lagny, I pray you, do more honor to my poor house than your young friend." Now, it must be confessed, the good old lady was hungry; and hunger is an overruling passion. The duke helped her to food and wine, and then, having done what second thoughts had shown him was only courteous to a lady, he turned, under the influence of the same better thoughts, toward Edward, who was still talking in a whisper to Lucette, while she, on her part, could hardly answer a word for weeping. "Young gentleman," said De Rohan, holding out his hand, "do not let us part bad friends. Remember, first, that if there be any validity in this marriage it is always better to keep well with a wife's relatives; and, secondly, that one of my house, above all others, may well feel mortified and enraged at an alliance which under no circumstances we could have desired or sanctioned. Recollect our family motto, —'Roi ne puis; prince ne daigne: Rohan je suis;' and pride is not so bad a thing as you may think it now. If it be pride of a right kind, it keeps a man from a world of meannesses. As to this young lady, I will take care of her, and, now that my first fit of passion is past, will treat her kindly. Be sure of that, Lucette; for I have even got a notion, by some bad experience, that a portion of love is no evil in the cup of matrimony. However, the question of this marriage must be a matter of consultation between my brother Soubise and myself, and the lawyers too; for I will not conceal from either of you that
- 66. Soubise, who has more to do with the business than I have, will break it if he can." Edward took the proffered hand; but he only replied, "His Eminence the cardinal said that he had made it so fast there was no power on earth or in hell to break it. But that must be determined hereafter, my lord duke. At the end of two years I will claim my wife. In the mean time, where is Monsieur de Soubise?" "Go not near him! go not near him!" said De Rohan. "By my honor, there would be blood-shed soon! He is at Blavet, I fancy, now, on his way to England; but I will write to him this night, and, if possible, you shall have his answer at Nantes. You must not expect any thing very favorable to your pretensions; but, whatever it is, it shall be sent." "My lord, if I might ask one favor, I would do it," said Edward. "It is this. From what you have yourself said, and from what others have told me, I infer that Monsieur de Soubise is of no very placable nor temperate disposition. He himself has had some share in producing both what you look upon as a misfortune and what had nearly proved the destruction of Lucette and myself, by sending—with very good intentions, doubtless, but I think very unadvisedly—letters and other matters to the very residence of the court, which betrayed our coming to his Eminence the cardinal. Had that not been done, we should in all probability have passed without question, and I should have been able to restore this dear girl to her relations as Mademoiselle de Mirepoix. As it is, my wife she is and must remain; but I would rather that she was under your care than that of the prince, for she has this evening suffered too much for an event, which she could not avoid without dooming herself and me to destruction; and I would fain that the same or perhaps more should not be inflicted upon her from another quarter. Lucette will explain to you much that I have not time to tell, for I see Madame de Lagny has risen, and it is growing so dark that I fear we must depart." "I can promise nothing," said the duke, "but that I will do my best."
- 67. Thus saying, he turned toward Madame de Lagny, who by this time had some lights on the table before her, and addressed to her all those ceremonious politenesses which no one knew better how to display, when not moved by passion, than the Duc de Rohan. In the mean time, Edward and Lucette remained at the darker side of the room; but, had it been the broadest daylight, their natural feelings would have suffered little restraint. The contrast of Edward's love and tenderness with the cold harshness of her own relations made all her affections cling closer round him than ever, and she hung upon his breast and mingled kisses with his, while the tears covered her cheeks and sobs interrupted her words. "Oh, Edward," she said, "I wish to Heaven that I were indeed but the grandchild of good Clement Tournon, of Rochelle, as you once thought me! We might be very happy then." Mingled with his words of politeness to Madame de Lagny, the duke had been giving some orders to his own attendants; and at length he said, "Now, young gentleman, it is time to depart. Madame is ready." One last, long embrace, and Edward advanced to the side of the duke. He did not venture to look at Lucette again, but followed Rohan and Madame de Lagny closely into the outer hall, thence through a small court and a place d'armes, in each of which were a number of soldiers fully armed, and then by a covered way to the water-gate, to which point the small boat had by this time been brought round. There was still a faint light upon the river; but a lantern had been placed lighted in the bow of the boat, and in a few minutes the old lady and her young companion were landed on the other side. One of the boatmen lighted them up to the carriage, and Edward, after bestowing a piece of money upon the man, took his seat beside Madame de Lagny, who gave orders to proceed toward Nantes, stopping, however, at the first auberge where any thing like tolerable accommodation could be found.
- 68. "Ah, poor Monsieur de Rohan!" she said, with perhaps not the most compassionate feelings in the world. "He is much to be pitied; and, indeed, he ought to feel, as he said, that some love in marriage is a very good ingredient. He ought to know it by experience; for his own good-for-nothing dame cares not, and never did care, for him; and it is the common phrase in Paris that she has so large a heart she can find room in it for everybody except her husband. Why, I know at least ten lovers she has had besides the Duc de Candale, who is more her slave than her lover, and who"—— Just at that moment, the horses having been put to, the coachman gave a sharp crack of his the whip, the coach a tremendous jolt, and Madame de Lagny brought her story to an end, somewhat to the relief of her young companion.
- 69. CHAPTER XXII. For the first time in life—and it was very early to begin—Edward Langdale felt that loneliness of heart which parting for an indefinite time from one we dearly love produces in all but the very light or the very hard. He had never loved before; he had never even thought of love; but now he loved truly and well. He might not indeed have loved even now, for he and Lucette were both so young that the idea might not have come into the mind of either; but their love had been a growth rather than a passion; and, as the reader skilled in such mysteries must have seen, it had been watered and trained and nourished by all those accidents which raise affection from a small germ to a beautiful flower. First, she had nursed him so tenderly that he could not but feel grateful to her; then she had been cast upon his care in dangers and difficulties of many kinds, so that deep interest in her had sprung up. Then, again, she was so beautiful, in her first fresh youth, that he could not but admire what he protected and cherished. Then she was so innocent, so gentle, so ductile, and yet so good in every thought, that he could not but esteem and reverence what he admired. Then had come his turn of nursing, and the interest became warmer, more tender; and at length, when the mere thought of stating, in order to account for their companionship, that they sought to be married first entered the mind of each, it let a world of light into their hearts, and the whole was pointed, directed, confirmed, by the sudden ceremony which bound them together. They had promised at the altar to love each other forever, and they felt that they could keep their word. But Edward, as he rolled along by the side of Madame de Lagny, could not help asking himself painful questions: "I shall love her ever," he said to himself; "but she is so young, so very young,—a
- 70. mere child! Will her love last through a long separation? will not her feelings change with changing years? does she even love me now as I love her?" Luckily he asked himself the last question, for it went some way to answer the others to his satisfaction. There had been something in her embrace, in her kiss, in her eyes, in her clinging tenderness, which told him that she did love as he did; and he, feeling, or at least believing, that he would love still, however long they might be separated, learned to credit the sweet tale of Hope and believe that she would love constantly too. Nevertheless, he felt very sad; and yet he exerted himself eagerly and successfully to make the journey pass as pleasantly as he could to poor Madame de Lagny, who, though she had not undertaken her disagreeable task out of any affection to either Edward or Lucette, but merely in obedience to the wishes of Richelieu, had learned to love both her young companions, and had taken their part sincerely in the discussion with the Duc de Rohan. She was both a keen- sighted and a clear-minded old lady; and she saw well the gloomy sadness of Edward Langdale, and understood its cause; but she saw likewise that he was making every effort to show her courteous attention; and no old women are ever ungrateful for the attention of young men. For three days the weary journey back to Nantes continued; and in that time the good marquise contrived to store the young Englishman's mind with many of her own peculiar apothegms, some good and some indifferent, but all the fruit of much worldly experience grafted upon a keen and sensible mind. "Never despair, my son," she said. "Many a man is lighted on his way by a candle; nobody by a stone. Of a misfortune you can remove, think as much as you like; of a situation you cannot change, think as little as possible. If you have a marsh to go through, gallop as fast as you can; and, if you have a heavy hour, fill it with action. A wasp
- 71. will not sting you if you do not touch it; and we do not feel sorrow when we do not think of it." Such were a few of the old lady's maxims, and one of them struck Edward Langdale's fancy very much. "If you have a marsh to go through," he repeated to himself, "gallop as fast as you can; and, if you have a heavy hour, fill it with action." He thought that the next two years would indeed be a marsh to him, and he resolved to gallop through them as fast as he could. But there was one sad reflection which he could not banish, one point in his situation which gave him anxiety rather than pain. He knew not how to hold any communication with his young bride. He was well aware that every effort would be made to prevent it. Lucette had been once sent to England to keep her out of the hands of the Duchesse de Chevreuse: where might she not be sent now? Her two cousins Soubise and Rohan were constantly roving from place to place, and there was as little chance of any letter from him finding her as of any news of where she was reaching him. The old fable of Midas telling his misfortune to the reeds is founded upon a deep knowledge of human nature. Man must have some one to share the burden of heavy thoughts, and Edward told his to Madame de Lagny. The old lady was better than the reeds, for she whispered consolation. "I can help you but little, my son," she said; "but, if you could attach yourself to the cardinal, he could help you a great deal. However, I will do the best I can for you and the dear child your little wife. If you want to write to her, send your letter to me at the court, wherever it is, and the letter shall reach her sooner or later. I will find means to let her know that she must send hers to me likewise, and they shall reach you; if you will keep me always informed of where you are." Edward not only pressed her hand, but kissed it; and not five minutes after, when they were within ten miles of the city of Nantes, a man came riding at full speed after the carriage, drew up his horse at the great leathern excrescence called the portière, and asked, in a brusque tone, if Monsieur Langdale was in the coach.
- 72. "Yes; I am he," answered Edward. "What want you with me?" "A letter," replied the man. And, handing in a sealed packet, he turned his horse's head and rode away. It was still early in the day, and the youth, breaking open the letter, read the contents. They ran thus:— "My Lord and Brother:— "On the wing for England, I have received your letter. Tell the insolent varlet that he shall never see her face again, the devil and the pope and the cardinal to boot, or my name is not "Soubise." Edward's brow became fearfully contracted, and he muttered, "At the end of the earth." "Show it to me! show it to me!" exclaimed Madame de Lagny, who was not without her share of woman's curiosity. "What is it makes you look so angry, my son?" Edward handed her the letter, and she read it with attention, but not with the indignation he expected to see. On the contrary, she seemed pleased and amused. "Let me keep this," she said. "Methinks that Monsieur de Soubise may find the triple alliance of the devil, the pope, and the cardinal to boot somewhat too much for him. The cardinal alone might be enough, without two such powerful auxiliaries. But let me keep it. It can be of no value to you." "Oh, none!" answered Edward. "Keep it if you will, madame. But the Prince de Soubise shall find that, if he have a strong will, I have a strong will also; and, if he have some advantages, we have youth and activity and resolution." "And the Cardinal de Richelieu," said Madame de Lagny, emphatically: "he is not the man to leave any work incomplete, nor
- 73. to be bearded by any one. However, we must be near Nantes by this time. Now let us consider what your course is to be when we arrive." The good marquise then proceeded to indoctrinate her young companion with all the forms of a court, which, though not so rigid as they afterward became,—for Louis XIV. was the father of etiquette,—were sufficiently numerous and arbitrary to puzzle a young man like Edward. He found that, although he had once by the force of circumstances won easy access to the cardinal prime minister, he had now various ceremonies to go through before he could hope for an audience. To call, to put down his name and address in a book, to see principal and secondary officers, and to give as it were an abstract of his business, were all proceedings absolutely necessary, Madame de Lagny thought, before he could see the cardinal; and Edward, with a faint smile, asked her if she did not think it would be better for him to commit a little treason as the shortest way to the minister's presence. "Heaven forbid!" cried the old lady. "But in the mean time you must go to an auberge near the chateau, where his Eminence can find you at any moment." And she proceeded to recommend the house of an excellent man, who had been cook to poor Monsieur de Lagny, and now, she assured Edward, kept the very best auberge in Nantes. At length the city was reached, and the coach drove straight to the castle, where Madame de Lagny took a really affectionate leave of Edward and retired to her own apartments. The young Englishman then proceeded to inquire for Richelieu, found he was absent at a small distance from the town, and, having written his name in a book, betook himself to the inn which his travelling-companion had mentioned. In the court of the castle he had seen no one but a guard or two and some servants at the door of the hall. In the great place there was hardly a human being to be seen,—no gay cavaliers on horseback or on foot, no heavy carrosse with its crowd of laquais. At the other side of the square, indeed, near the end of the little street which led toward the dwelling of Monsieur de Tronson, was a group of workmen; and another larger group just appeared beyond
- 74. some buildings close by the river-side. But, altogether, the whole town had a melancholy and deserted look. A sort of ominous silence reigned around, too, which Edward felt to be very depressing to the spirits, especially in a country celebrated even then for the light hilarity of its population. The inn, however, was fresh-looking and clean, and the landlord, who soon appeared, although he was not at the entrance as usual when the coach stopped, was the perfection of a French aubergist,— as polished as a prince, and full of smiles. While Pierrot la Grange and Jacques Beaupré stayed by the carriage, at their master's desire, to take out the little sum of his baggage and to bestow a small gratuity upon the coachman, the host led his guest up to a large, somewhat gloomy chamber floored with polished tiles, recommended his fish—the best in the world—and his poultry, which he asseverated strongly were the genuine production of Maine, and took the young gentleman's pleasure as to his dinner. He had hardly gone when the two servants appeared, bringing various articles; but their principal load was evidently in the mind. The face of Pierrot, which temperate habits had not yet improved in fatness, though it had become somewhat blanched in hue, was at least three inches longer since they entered Nantes; and Jacques Beaupré, always solemn even in the midst of his fun, was now not only solemn, but gloomy. "I wish we were safe out of this place, sir," said Pierrot, shutting the door after him. "It is a horrible place!" "What is the matter?" asked Edward: "the whole town looks sad, and you both seem to have caught the infection." "Did not the landlord tell you, sir?" said Jacques Beaupré. "I thought landlords always told all they knew, and a little more. But I suppose he has lived long enough near a court to keep his tongue in his mouth, for fear somebody should cut it out."
- 75. "The matter, sir, is this," said Pierrot: "the poor young Count de Chalais, who was confined in the dungeons close under the room where they put you, has been condemned to die this morning,—they say, for a few light words." "Indeed!" said Edward, with a somewhat sickening memory of the dangers he himself had seen: "that is very sad. But probably the king will pardon him." "Oh, not he," answered Pierrot: "they say the poor countess, his mother, has moved heaven and earth to save him, without the least effect. His head is probably off by this time." "No, no; that cannot be," rejoined Jacques: "did not the boy tell us that the two executioners had both been spirited away?" "Yes, but he said that a soldier—a prisoner—had been found to undertake the job," answered Pierrot. "Oh, it is a bad business, Master Ned! They say the queen herself has been brought before the council, and the Duke of Anjou threatened with death, and half the court exiled, and the cardinal in such a humor that——" "That every one as he walks along is feeling his ears, to be sure that there is any head upon his shoulders," added Jacques Beaupré. "Would it not be better for you, sir, to go to that good Monsieur de Tronson, and be civil to him, and make as many friends as possible?" Edward paused in thought for a moment, and then replied, "That is well bethought, Beaupré; for though I think I have nothing to fear, yet in common courtesy I owe my second visit to one who has been so kind to me. I will go directly. Let the landlord know that I may be a little later than I mentioned at dinner." Edward put on his hat and went out into the place, taking care to mark particularly the position of the auberge, that he might not be forced to inquire his way in a town where so many dangers lurked on every side. The road to Monsieur de Tronson's house was easy; and, crossing the square, the young gentleman directed his course
- 76. toward the end of the street where, when passing in the coach, he had seen a crowd of workmen, who were still gathered round a spot about a hundred and fifty or two hundred yards in advance. On approaching nearer, Edward caught sight of a platform of wood raised some eight or ten steps from the ground. He could only discern a part, for the people had gathered thickly round; but, though he had never before seen the preparations for a public execution, it flashed through his mind at once that this was the scaffold on which the unhappy Chalais was to suffer. To avoid the terrible scene, he turned toward the left; but, just as he was approaching the end of the street, a shout came up from the water- side and a dull rushing sound from the southeast. A large crowd poured into the square from both sides; and before Edward could escape he was caught by the two currents and forced along to within thirty yards of the scaffold. He tried to free himself and force his way out, but a warning voice sounded in his ear. "Be quiet, young gentleman," said an elderly man close by, speaking in a low tone. "This young count has to die, and, if he be your best friend, take no notice. Suspicion is as good as proof here just now. Look where he comes!" Edward turned his eyes in the direction to which the old man was looking, and beheld a sight which was but a mere prologue to the horrors that were to follow, but which could never be banished from his memory. Surrounded by a body of guards came a tall, handsome young man, without his cloak, as if he had been torn from his dungeon unprepared, but still showing, in such habiliments as he did wear, all the extravagant splendor of the times. By his side, with her hand passed through his arm, as if to support him, and pouring a torrent of words into his ear, was an elderly lady in a widow's dress. Her face and carriage were noble and dignified, though lines of past grief and present anguish were strongly marked upon her countenance; but when she lifted her eyes toward the scaffold, and beheld there a stout, bad-looking man leaning on a large, heavy sword, a sort of spasm passed over her features.
- 77. "That is his mother," whispered the same voice which Edward had heard before. Behind the mother and the son came the confessor, a dull-faced, heavy monk; and then a good number of guards, and one or two men in black robes,—probably exempts, or other inferior officers of the court. But the eyes of Edward Langdale were fixed upon the mother and her son; and the thought of his own dear mother gave him the power—I might almost call it the faculty—of sympathizing with the noble-minded woman, to a degree that made the whole scene one of actual agony. "I wish I could get out," he said, speaking to the old man, who was jammed up against him: "this is horrible. Can you not make way?" "Try to force your way through the castle-wall," replied the other, cynically: "you have but to see a man die, young gentleman." "Ay, but how?" said Edward. "By the sword," said the old man: "it is an interesting sight,—much better than by the cord. I have seen every execution that has taken place in the city for twenty years. Perhaps I may see yours some day. They are fine sights,—the only sights that interest me now; but this is likely to be a bungled business, for the old countess there bribed both the executioners to get out of the way, and this fellow does not understand the trade. He is paler than the criminal. See how he shakes!" Edward raised his eyes for an instant and saw the unhappy mother supporting her luckless son up the very steps of the scaffold,—not that he wanted aid, for his step was firm and his look bold and frowning. There was a fearful sort of fascination in the sight; and the lad gazed on till he saw the last embrace taken and the young count make a sign and speak a word to the executioner. Then he withdrew his eyes, till, a moment after, there was a shrill cry of anguish and a murmur amongst the crowd; and he looked up again only to see the
- 78. wretched young man, all bleeding, leaning his wounded head upon his mother's bosom. The executioner had missed his stroke. Again and again he missed it. He complained of the sword: a heavier one was handed up to him; but still his shaking arm refused to perform its hideous office, till, after more than thirty blows,[4] the head of the unhappy young man was literally hacked off, almost at his mother's feet. The noble woman raised her hands and her eyes to heaven, exclaiming, "I thank thee, O God, that my son has died a martyr and not a criminal!" The last acts of the terrible drama Edward did not see. He felt as if his heart would burst with the mingled feelings of indignation and horror which all he had beheld awakened; and after the second or third blow he kept his eyes resolutely bent down, till the pressure of the crowd relaxed as the spectators of the bloody scene began to disperse. Then, sick at heart, and with a strange feeling of hatred for the world, he turned his steps back to the inn. He was in no mood for conversation with any one.
- 79. CHAPTER XXIII. It was eleven o'clock on the following day when Edward Langdale appeared at the door of Monsieur de Tronson. The laquais said he did not know whether his master was visible or not, but he would see; and, leaving the young Englishman in an ante-chamber, he went in and remained some five minutes. At his return he asked Edward to follow, and introduced him into the bed-chamber of the secretary, who welcomed him, he thought, rather coldly. "I hear, Monsieur de Langdale," said De Tronson, "that you have accurately fulfilled the injunctions of his Eminence and your word. That, my good cousin, Madame de Lagny, has told me; but I think you should have been here earlier." "It was my intention, sir," replied Edward, seating himself in a chair to which the secretary pointed, near that in which he himself sat, wrapped in a large dressing-gown, by the fire, though it was the month of July. "After having left my name in the ante-chamber of his Eminence, I went to my auberge for a few minutes, and then came out, with the intention of paying my respects to you; but I was stopped by a great crowd of people and forced to witness a dreadful scene, which rendered me incapable of holding any rational conversation with any one." "Ha! you were there!" exclaimed the secretary, suddenly roused from the sort of listless mood in which he seemed plunged when Edward entered. "What happened? Tell me all. But first shut that door, if you please. I am ill, or I would not trouble you; but it is well to have no listening ears in this place, whatever one has to say."
- 80. Welcome to our website – the perfect destination for book lovers and knowledge seekers. We believe that every book holds a new world, offering opportunities for learning, discovery, and personal growth. That’s why we are dedicated to bringing you a diverse collection of books, ranging from classic literature and specialized publications to self-development guides and children's books. More than just a book-buying platform, we strive to be a bridge connecting you with timeless cultural and intellectual values. With an elegant, user-friendly interface and a smart search system, you can quickly find the books that best suit your interests. Additionally, our special promotions and home delivery services help you save time and fully enjoy the joy of reading. Join us on a journey of knowledge exploration, passion nurturing, and personal growth every day! ebookbell.com