校事
红茶是发酵茶吗
Copyright © 2000, Idea Group Publishing. Copying without written permission of Idea Group Publishing is prohibited.
2 Petrides, Khanuja-Dhall & Reguerin
C o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g C
o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g CASE NARRATIVE
Background
Midwestern University (MU) has an enrollment of approximately 15,000 students.Since it was founded, the mission of MU has been to provide world-class leadership in teaching and rearch. Within MU there are 15 academic departments and veral administrative units. University administration had historically taken a very centralized approach to program enrollment, recruitment, financial aid, and general administration of student-related matters. However, more recently, top-level administration has encouraged individual departments to take more local control of their planni
ng, ranging from student administration to budget tting. The push for local or departmental control has not been accompanied by the requisite development of reliable information systems necessary for both short- and long-term planning. This decentralized approach has placed departments at a distinct disadvantage due to increasing levels of account-ability at the department level.
Historically, information such as student enrollments and financial aid allocation flowed downward from central administration offices to the departmental level. The upward flow of information consisted of a t of checks and balances associated with departmental graduation requirements. In addition, data that were specific to the department level did not flow upward (e.g., faculty advising lists and student progress reports). Administrative divisions were centrally managed with multiple databas tracking data in functional units. For example, enrollment data were maintained and controlled by admissions, but the graduate studies office controlled doctoral student data. Many of the systems were run with old and outdated software, and the university struggled with the lack of a coordinated information system that managed all data collected throughout the university. This resulted in issues of data integrity, redundancy, and accuracy, with a low level of trust concerning the interpretation of data.
Enrollment data were maintained at the university level. The data were available to assist the dep
artment in knowing how many students were enrolled during a particular mester. However, it could take three to four weeks to obtain data from the central student information system, and field definitions were ldom defined. Additionally, becau students were not centrally tracked through the various stages of doctoral completion, it was difficult if not impossible to ascertain the types of class, rvices, and faculty commitment that students required with any degree of certainty. Departments relied on anecdotal information to conduct planning, and this became a standard and acceptable practice by default. Additionally, many faculty suspected that there were dozens of students who slipped through the cracks in the process somewhere along the line and might have been precipitously clo to dropping out.
t5000
There was also a high level of dissatisfaction among MU students with regard to information management. Students were frustrated with the number of repetitive steps and process involved in their educational experience. For example, students needed to register for class at the registrar’s office. However, depending on the class students wanted to register for, they may have needed to receive departmental signatures prior to registration and then go to an entirely different office to make tuition payments. Becau of the amount of time spent in completing the tasks, students’ frustration level only incread when the data across the areas could not be shared.
The Arts and Humanities (A&H) department has approximately 200 doctoral graduate students, 200 graduate master students, and 300 undergraduates enrolled. Unlike the master家庭教育的论文
The Politics of Information Management 3
C o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g C
o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g and undergraduate students, who have structured two- and four-year programs, doctoral students went through veral different stages of enrollment; first as graduate students enrolled in class,then as doctoral candidates once they pasd comprehensive exams, followed by a period of time during which they took independent disrtation-related methods cours and disrtation writing minars. This multi-stage process was very complicated to track and the department had been unable to determine with much accuracy at what stage in the matriculation process their 200-plus doctoral students were at any given time.
This had many implications for departmental planning. The opportunities and challenges prented by a more decentralized structure of decision-making needed to be supported by reliable information.
In conjunction with this challenge, the department began to conduct long-term planning for doctoral cour offerings and faculty disrtation loads. This affected planning for core cours, rearch minars, and disrtation writing workshops.
Additionally, there were implications for faculty workload since work with doctoral students could be a very time-consuming process at various stages of their degree. In fact,the proposal and final writing stage for doctoral students working on their disrtations often required a large investment of faculty time, mainly consisting of reading draft chapters and supplying timely feedback.
The Politics of Information Sharing
With the University’s push to a decentralized model of operation, departmental accountability and ownership of doctoral student data were becoming a priority. The need for the department to track and asss doctoral student status was crucial to both the doctoral students’ and departments’ success. Members of the department decided that they needed to do something about the situation. They agreed to tackle their first goal – how to improve access to student information.
In an attempt to address this issue effectively, a needs asssment was conducted. This consisted of determining what type of information was required about doctoral students in order to do more sho
rt- and long-term planning. During the planning process, the depart-ment faculty realized they did not even know how many doctoral students had continuous enrollment over the past two mesters, let alone how many students were projected to graduate that year. There were larger issues of completion and attrition that faculty wondered about but emed afraid to find out. Simple questions were unable to be answered,such as: how long do doctoral students take to complete the program, how many students have completed their courwork but not yet taken their comprehensive exams, how many students need to take a disrtation writing minar the next mester, and how much financial aid support do students need to graduate.
番禺景点
狂草书法Not only were there student-related questions without answers, but there were also issues of faculty workload. There were 25 full-time faculty members in the A&H depart-ment. Seven of them were untenured but on the tenure track. It had been brought to the Dean’s attention in promotion and tenure reviews that the junior faculty might have a disproportionate amount of the doctoral student load. However, when asked, the department chair was only able to answer the question bad on general estimates and hearsay. There were no reliable data regarding faculty workload issues. This lack of information regarding doctoral students and faculty workload only made stronger the department’s chair request that the information management of the department be improved upon.
The departmental culture was one in which information was heavily protected.Traditionally, the sharing of information had been the source of political disputes. Faculty
戴尔r7204 Petrides, Khanuja-Dhall & Reguerin
C o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g C
o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g neither felt that they gained anything by sharing information about doctoral workload, nor did they e the need to. In this ca, nior faculty members typically had a lighter doctoral student workload than junior members and wanted to avoid workload reallocation. How-ever, junior faculty who had a heavy workload struggled to obtain and share doctoral student information with other faculty. In this ca, the issues only added to the clod nature of sharing information in the department, since information sharing behavior was neither recognized nor rewarded.
Whether we like it or not, information politics involves competing interests,disnsion, petty squabbles over scare resources (Davenport, 1997, p. 78).
A First Step
Two years earlier, the department chair had instructed his administrative assistant to begin to collect and maintain departmental doctoral student data using a Lotus spreadsheet.The data were kept independently of the university-wide information systems. Numerous challenges associated with creating, sharing, and updating the spreadsheet files were faced.The historical operation of the department was heavily reliant on another office’s data, and faculty’s lf-management of their doctoral students led to information that was not readily available at the departmental level. Furthermore, it was very difficult for the administrative assistant to consolidate the information from the disparate systems and faculty members.Specifically, the data that were to be compiled included information such as: the number of credits for students currently enrolled, their year in the program, their comprehensive exam completion status, their faculty member advir, and the amount of time students had left to finish their courwork.
As indicated, this information was not centrally located and each system varied in type and form. Within the department, some data were in hardcopy only, filed in a file cabinet or in handwritten notebooks that faculty ud for personal tracking of their students. Some of the information was not even documented or available in an accessible system. With so many varying types of systems and the data being scattered throughout the department, the effort to consolidate the information into a s
preadsheet was difficult. In order to create a workable tool, the scope of the data collection effort was limited only to departmental doctoral student information.
Once the information was collected and consolidated into the spreadsheet, reports were summarily disregarded by faculty. When looking cloly at why the spreadsheet failed,veral items were identified. For example, there was the limitation that spreadsheets impo on data – data must be depicted in columns and rows, and the ability to crosscut data is limited. For example, a header row contained student year, faculty advir, and the number of years that student had been enrolled. The spreadsheet had 50 columns across and more than 200 rows down. Becau a spreadsheet cannot be queried, the only way to find or organize the information was by sorting the entire spreadsheet. This became cumbersome becau, if a multiple column sort was conducted, Lotus would sort one column at a time,independent of the other columns, with the end result being a sorted list of all students not just the category desired. The administrative assistant tried to counteract this by taking a portion of the complete spreadsheet and cutting and pasting it into another file. This resulted in multiple spreadsheets with information that needed to be updated in eight or nine different files. Even if the person responsible for doing this kept track of the updates, it would be extremely inefficient, redundant, and prone to error.
The Politics of Information Management 5
陈露
C o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g C
o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g The inability to develop special views of the data and custom reports was a limiting factor with the spreadsheet. This querying limitation only incread the lack of support and u for reliable information. A cond, and more obvious challenge, was the administrative assistant’s lack of sophistication and training around the software itlf.
The Web-Bad Relational Databa Project
Despite the initial failure, the chair of the department asked two technologically minded faculty members, both untenured, to write a proposal for building a relational Web-bad databa that would consolidate and centralize data from veral different areas of the university, including other administrative offices outside the immediate department. They submitted a proposal to build a Web-bad, password-protected databa that would be accessible to all faculty. The propod system would be easy to u; they estimated that it would take approximately two hours to train a computer-
knowledgeable individual to u the system. The data would reside in one file, and reports could be created automatically.They would provide two-hour training for the administrative assistant, a two-page list of instructions of how to import data and produce reports, and a one-page list of instructions for faculty members on how to access and u the databa via the Web. They estimated that it would take them eight months to complete the project. The department chair gave them $7,000 the next week to begin their work.
The design team was led by the two faculty members. An outside consultant who specialized in databa design was hired to join the team. Becau the Web-bad technology was somewhat new to the department, a consultant specializing in Web development was also brought on to help create the proposal and pilot system.
In creating a proposal that would define the scope of the project, the resources required,and the required information for the databa, the two faculty members divided the project into three main phas – planning, design, and implementation. This provided them with a framework that gave measurable and clear checkpoints that were dependent on departmen-tal faculty approval.
The planning stage first involved a requirement study that consisted of identifying a comprehensive li
st of the department’s information needs. This also required looking at external data requirements and the systems that data would come from. The additional data that would be gathered from across the university included data from the Student Informa-tion System (SIS) managed by the Registrar’s Office, the Doctoral Student Databa (DSD)managed by the Graduate Studies Office, and the Student Payment System (SPS) managed by the Student Accounting Office. Student data for each of the systems were to be consolidated into the A&H relational Web-bad databa, along with additional data that were collected at the departmental level only (e.g., faculty advirs and disrtation chairs).
The two faculty leaders conducted interviews with each of the faculty and prioritized requests from the departmental members and the chair. The need for new data that had not been collected previously by any office was also identified. The compilation of all the requested data came from approximately 20 different subsystems both manual and electronic. As described earlier, the systems ranged from word processing to handwritten notebooks.
The next pha required designing the relationships between the data elements and tables.The databa consultant helped to incorporate a databa design that was able to depict the relationships between each of the different data tables with relative ea. This provided an initial und
erstanding of system complexity by focusing on the relationships between data, data types,and source. This exerci was esntial in proactively understanding how the new system would
6 Petrides, Khanuja-Dhall & Reguerin
g C o p y r i g h t I d e a G r o i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g C o p y r i g h t I d e a G r o u p P u b l i s h i n g be queried and what information would be collected in the new system. Diagram 1 shows a relational schematic of how a few tables in the databa would be linked by student social curity number, a primary and unique key across each table.
The diagram illustrates the relationship between the new tables to be created and the source of the data. The design team determined that approximately 32 tables with 500 data
elements would be required in the new Web-bad system. This included information such as: demographics, address, first enrolled, last attended, disrtation chair, whether students attended school full time or part time, and when their doctoral candidacy expired.
The issue of data maintenance was raid as a main concern in the design pha, and the team recommended a system manager to keep the data integrity at an optimal level. The team lected software tools bad on the data complexity and faculty interviews. Having a clear understanding of the faculty requirements concerning doctoral student information,explicitly outlining the data relationships, and asssing the current mix of systems and interfaces, the team was able to confidently lect effective software tools for implementa-tion. The main goal surrounding the lection process was to identify a ur-friendly and intuitive front-end that would provide faculty with ea and functionality for sharing and accessing data.
The last pha, implementation, consisted of running a pilot with faculty, training the faculty, and receiving sign-off approval from the chair to operationalize the entire system.In piloting the system th
e two faculty members demonstrated the capability of the new system at a faculty meeting and also provided one-on-one demos. Bad on the demos,faculty members requested even more features and functionality from the system. Not only did the team implement the requested functionality, but they also incorporated an automated feedback form that would allow new feature requests to be delivered to the core develop-ment team on an ongoing basis. For example, if a faculty member identified a new feature she or he wanted, the faculty member could complete an online form that would forward the request to the right development team member. In addition, a respon could then be provided back to the faculty member indicating when and if the propod feature would be integrated. Up to this point, the core team thought the support for the system was mostly positive and energetic.t F a c u l t y e c u r i t y N u F a c u l t y M F a c u l t y M e w D a t e T a b l e m b e r a t e a m P a s s e o n D a t e Diagram 1: Sample Relational Table
