Decision With Data Warehouse For Higher Education

Published Date: 02 Nov 2017

Abstract

Data warehousing has been developed to meet a growing demand for management information and analysis that could not be met by operational systems, but many administrative people and executives are a little unclear whether higher education institutions need it, how it works and how it benefits. It is likely that once they fully understand exactly what a data warehouse can do, they will decide that one is needed. In challenging times good decision-making becomes critical. The best decisions are made when all the relevant data available is taken into consideration. An excellent source for that data is a well-designed data warehouse. The goal of this research is to explore why the university do not adopt readily the use of a Data Warehouse, what changes could be made, and how other universities could benefit from this research into Data Warehousing for higher education institutions, so the university can improve its decision making process.

Table of Contents

Abstract

Acknowledgements

Table of Contents

List of Figures

List of Tables

Chapter 1 – Introduction

Thesis Statement

Problem Statement

Goals of the project

Method

Scope of Project

Chapter 2 – Review of Data Warehouse Technology

2.1 Literature overview

2.2 What is Data Warehouse?

2.3 What is Business Intelligence?

2.4 Relationship of BI, data warehousing, OLAP, and Analysis Services

Chapter 3 – Project Methodology

Chapter 4 – Lessoned Learned

References

Appendix A - Interview Questionnaires

Appendix B – Survey Result

Chapter 1

Introduction

Thesis Statement

While the business organizations recognize benefits of data warehousing for decision making, it is not used widely in higher education. There has been remarkably little research being reported dealing with the data warehouse in higher education institutions. Especially, there is little research about successfully implementing data warehouse in this environment and about the benefits that they bring to such institutions. How can data warehouse improve the process of decision making in higher education?

Problem Statement

ABC University implemented administrative software package known as Colleague and Advanced Colleague, which was developed by Datatel. Colleague system is for a repository of student information and financial aid, and Advanced Colleague system for a repository of alumni and fund raising activities. The Colleague and Advance Colleague window application collect, distribute, and enable collaboration on student records, alumni information, and fund raising activities. The Unidata database system from IBM stores data from the Colleague and Advanced Colleague windows applications. The university has continually upgraded the Colleague student information systems over the years. However, other than software version upgrades, the modification of existing applications is extremely inflexible for customization because of its system architecture. The ABC University is already aware of the problem with the Colleague system, but there is not enough funding for the university to move to a new system. The main problem with the Colleague system is producing ad-hoc reports. Due to the Colleague system architecture, it is difficult for end users to access operational data. Building new reports to answer business questions is a slow and expensive process because Unidata used by Colleague system is not relational database. To create a new relationship between files, an expert’s programming skill is required. ABC University does not have a on-site colleague programmer and hiring a Colleague programmer is expensive. To overcome the problem, ABC University implemented Data Warehouse built on SQL server. Although the development of the data warehouse system for ABC higher education institutions started 6 years ago, it is still used for operational activities in limited departments and rarely used for strategic purposes. This research investigates the use of data warehousing in ABC University, documents its value, and identifies the reasons of the lack of adoption of data warehousing and business intelligence in the educational environment.

Goals of the project

This research identifies why ABC university staffs do not adopt readily the use of a Data Warehouse. This research also explains the capability of general data warehouse and data warehouse architecture of ABC University. The primary benefit of using data warehouse is that it stores and provides information in such a way that it allows university executives to make critical decisions (Kimball, 2008). While the business world recognized benefits of data warehousing for decision making, data warehousing for decision making is not used extensively in ABC University. College administrative staffs need easy access to student information in order to serve students quickly. They also need easy access to alumni information in order to increase the fundraising activities. Data delivery tools, such as Web, Excel, Mail merge, Access, etc., placed on client workstations should allow staff to extract data from warehouse easily for local data analysis and reporting. The goal of this research is to understand how to encourage data warehousing and business intelligence in the educational environment

Method

The research on data warehouse for higher education is based on published research papers, case studies, and books on data warehouse architecture and technologies. Information on the subject will be also gathered from white papers, data sheets, product specifications, technology tutorials, implementation guides, and training materials provided on web sites by database software vendors, such as Microsoft, Oracle, IBM, Datatel, etc. Data collected will be generated based on relevant search categories. Relevant information will be selected and classified into different categories for analysis and comparison that could be used to support the project goal and objective.

For the purpose of finding resources on the research topics, the author will use the Internet search engines and libraries. The Internet search engines include Google Scholar, Google, MSN, and Yahoo search engines. The search will be conducted based on keywords. This technique is valuable as the author managed to retrieve useful literatures/information on the research topics.

The libraries used for the purpose of research include Dayton Memorial Library (DML) from Regis University and Denver Public Library (DPL). The Lumen library’s online catalog from Regis University will be used to find articles and books on the topics of research.

Personal experience in data warehouse design, enhancement, maintenance, administration, and business intelligence development will also contribute to this research subject. Additionally, the research will include the use of questionnaire and feedback from interviews from the decision makers in various departments.

The research will involve a series of interviews with staffs from the university who can identify reasons of resistance to using data warehousing technology. As an interview method, CoursEvalâ„¢, a web-based software product, will be used. CoursEvalâ„¢ allows interviewees to participate in survey questions online with easy access and the analyzed survey result will be delivered to a interviewer. The research will conclude with the results from the exclusive interviews and data analysis from personal experience in hopes that decision makers in ABC University realize Data Warehouse architecture can enhance their decision making process.

Scope of Project

This project concentrated on the reasons of lack of adaptability of a data warehouse in ABC University and on the study of the characteristics of a data warehouse technology that could be applied to design and build database system architecture for fast and reliable decision making tool.

Chapter 2

Review of Data Warehouse Technology

Literature Review

A data warehouse is a system that holds data consolidated from the source systems into central data storage. It typically keeps from the most current data to years of history and is queried for analytical activities or business (Rainardi, 2008). Since 1970, Data Warehouse (DW) has played significant role in strategic decision making for business, but universities were almost not listed in the major users of data warehouses until the early 2000s (Wierschem, McMillen, & McBroom, 2003). Higher education’s acceptance and adoption of data warehousing is low for decision making. Guan, Nunez, and Welsh (2002) states that only a portion of colleges' and universities' data are collected, processed and stored in their information system. As a result, academic rectors and deans complain about the lack of reliable and valid information about their fund raising, finances, students and personnel. A lot of information was collected but not centrally gathered making retrieval difficult. As stated colleges and institution tended to shy away in earlier data warehouse systems that were abit complex and required a lot of capacity to put up and were also rather expensive. The data collection system were spread and not centrally stored, for this reason data was unrealiable and very hard to retrieve or draw conclusions from.

Carlo, et al. (2007) identify several benefits for universities that can be reached by developing an academic data warehouse. For instance, data warehousing provides a centralized source of information which can be accessible across different academic units to analyze problems in a timely manner, supply the data necessary for developing the Institution’s strategic plan, and enable an administrator to make better business decisions based on historical data available in legacy databases. It also provides the institution with a central repository for all its data eliminating the need for many database systems that may be impossible to maintain in the long run. This also allowed for easy retrieval of data and analysis of historical data to enable elements like comparisons, trends and summary statistics be made which could be use to provide information to enhance decision making. Maintenance of institution data will be realized as management is central resources not spread.

Turban, et al. (2004) defines decision making as "a process of choosing among alternatives courses of action for the purpose of attaining a goal or goals" (p.40). They also define a Decision Support System (DSS) as an interactive computerized information system that retrieve data from various sources, allows building their own models, and supports business and organizational decision-making (p.105). The reason of constructing a data warehouse is to acquire information upon which decisions can be made. If the information is not used to make decisions, there is no reason to collect it. In order to make business decisions, a well-managed data warehouse can improve an organization’s decision-making capabilities.

Goleman et al. (2002) recognize the role of data is crucial in making good decisions. "As decision is more strategic, however, criteria for success become more complex and subjective" (Bonabeau, 2003, p.120). The decision that today’s higher education’s administrators have to make are more complex. This is where technology has a critical role to play. As recognizing the need for reliable information in decision-making processes, today’s computer technology can process and analyze the overwhelming volume of information fast and turn them to knowledge that decision makers need.

The development of a data warehouse provides a centralized source of information. Companies that build data warehouse use Business Intelligence (BI) to extract data from various different systems. A Business Intelligence technology in the university can provide a broad knowledge about students, faculties, staffs, and alumni (Carlo, et al., 2007).

Technology based on data warehousing can provide assistance for making good decisions. This technology can be applied to decision making at all management staffs in a higher education institution. Data warehousing technology will provide the necessary assistance in ensuring data is summarized and analyzed or compared in valid and desirable ways to provide information that can be used to make sound decisions. Data collected over time for instance can be used to determine trends and tell whether a business is growing and provide the rate at which it is growing. Summary data will also be available. The data warehouse technology given its versatility and adaptability can be utilized to ensure data entered at the transaction level is summarized and analyzed to ensure managers can use the information gathered to assist in decision making.

What is Data Warehouse?

Definition of Data Warehouse

According to Bill Inmon’s classic definition, a data warehouse is a "subject oriented, integrated, non-volatile, time variant collection of data in support of management’s decisions" (Inmon, 1993, p29). The data that are stored in the data warehouse usually come from transactional databases across the enterprise, and can include transactional databases.

Ralph Kimball in 2002 added to Inmon’s term as "a copy of transaction data specifically structured for query and analysis." Data Warehouses combine data from all types of sources and have the characteristics of subject oriented, integrated, time-variant (has a time component), and nonvolatile (no data are deleted: everything is stored, time-stamped and logged) (Sakaguchi, 1997).

Figure 2.2.1: Data Warehousing Architecture

Note: From Readings in Database Systems, 4rd Edition, by Hellerstein, Joseph M., and Michael Stonebraker, 2005. The MIT Press.

As Bill Inmon (1993) termed the data warehouse (DW) as a "collection of integrated, subject-oriented data-bases designated to support the decision making process", the integration of this collection of sources is achieved through the use of ETL (Extract, Transform, and Load) processes.

ETL Processes

ETL processes are responsible for the extraction of the data from various sources, and in the processing of loading, data may have to transform to fit business needs, and ultimately loading it into a target data warehouse (Kimball, 2008). The data used in ETL processes can come from legacy and transaction database systems and transforming it into organized information in a user-friendly format to encourage data analysis and support fact-based business decision making (Kimball, 2008).

ETL (Extract, Transform, Load)

Once the data has been extracted and converted in the expected format, set of business rules is applied in transforming stage of ETL process. The data transformation may include various operations including but not limited to filtering, sorting, aggregating, joining data, cleaning data, generating calculated data based on existing values, validating data, etc (Kimball, 2008). The final ETL step involves loading the transformed data into the data warehouse.

There are a lot of ETL tools, including IBM DB2 Warehouse Manager, Oracle Warehouse Builder, and Microsoft with their SQL Server Integration Services.

Transactional Database vs. Data Warehouse

Even though applications with transactional database can perform routine data processing functions, transactional database is not able to do analytical processing, because the following reasons (Kimball, 2005):

Transactional databases contain only raw data, so the processing speed is significantly slower.

Transactional databases are designed for processing user requests such as retrieving data or updating the database, not designed for analysis such as queries, reports and analyses.

Repeating the same analysis on live data lead a different result from the previous run, because the data are being updated continuously.

Difference between operational database and data warehouses:

Aspects

Operational Databases

Data Warehouses

User

System Designer, System Administrator, Data Entry Clerk

Decision Maker, Knowledge Worker, Executives

Function

Daily Operations, Online Transaction Processing

Decision Support, Online Analytical processing

DB Design

Application Oriented

Subject oriented

Data

Current, up-to-date Atomic, Relational (Normalized), Isolated

Historical, Summarized, Multidimensional, Integrated

Usage

Repetitive, Routine

Ad hoc

Access

Read/Write Simple Transaction

Read mostly, Complex Query

System requirements

transaction Throughput, Data Consistency

Query Throughput, Data Accuracy

Figure 2.2.3: Operational Databases vs Data Warehouses

Note: Chaudhuri and Dayal, 1997; Inmon, 2005; Jarke et al., 2000; Kimball, 2005; and Mattison, 2006.

Data Warehouse is not a place to store duplicated data from operational database. It is important to be selective about what goes in a data warehouse. It is not wise to get every possible type of data from all the sources, just in case you might want to ask any question to your data warehouse (Jarke, Lenzerini, Vassiliou, & Vassiliadis, 2000).

The ultimate purpose of a data warehouse is to integrate enterprise wide organization data into a single repository from which users can easily extract data according to the organizational needs for the purpose of generating reports, performance analysis and decision making (Mattison, 2006).

Use of Data Warehouse

Bill Inmon defined a data warehouse is a collection of data that supports decision-making processes (2005). Traditionally, a data warehouse is constructed to collect and organize historical business data upon which decisions can be made. The organization that owns this information can analyze it in order to find historical patterns or connections that can allow them to make important business decisions (Inmon, 2005).

Until recently, only database experts were able to access to data in order to create the complicated queries necessary to retrieve, format, and summarize information for use by analysts and managerial decision makers. As lower levels of management are involved in the decision-making process more, the need for data warehouse has increased for direct end-user access to data warehouse data by people with limited database expertise.

Since a data warehouse can provide a single repository for all of data, its data is stable over time and DW allows sharing same data between different business units (Reeves, 2009). Data is loaded in scheduled time interval from the data providers.

The Data Warehouse structure is specifically designed for quick response time, which makes suitable as a reporting tool. Even reports requesting a large number of rows are usually returned in a few minutes (Reeves, 2009).

A Data Warehouse provides foundation and feedback information at the appropriate time through various business intelligence tools. A data warehouse is often a core component of a Business Intelligence infrastructure within an organization. BI is the tool to simplify information discovery and analysis as well as to delivery of useful information to the appropriate decision makers within the necessary timeframe to support effective decision making.

Business Intelligence Categories:

Type

Information seeking

Basic querying and reporting

What happened?

Business analysis (OLAP)

What might happen or What is something interesting?

Data Mining

What happened and why did it happen?

Dashboards and scorecards

Tell me everything, but don't make me work too hard.

Figure 2.2.5: Business Intelligence Process

Note: From Business Intelligence for the Enterprise by Biere, Mike, 2003. IBM Press.

Business inteligence

What is Business Intelligence?

It is important to acknowledge the need for good information in decision-making. Today’s information systems make the overwhelming volume of information and decision makers need tools to help analyze the complex situations that face them. Technology based on data warehousing can deliver the assistance that is needed for making good decisions at all management levels in an organization.

Kimball (2007) defines that Business intelligence (BI) can be any activity, tool, or process to get the best information to help the processing of making decision. In a data warehouse environment, BI refers to computer-based techniques used in identifying, uncovering, and analyzing business data (Kimball, 2007). BI technologies provide historical, current, and predictive views of business operations. BI can be considered as the types of reports and their respective audiences. Typical reports are Dashboard reports, Production reports, and Analytical reports (Biere, 2003).

Figure 2.3.1: Architecture of Business Intelligence

Note: From Business Intelligence for the Enterprise by Biere, Mike, 2003. IBM Press.

Dashboard reports

This is a highly summarized collection of key performance indicators using color and graphic to present business intelligence and decision support information for use by executives (Rasmussen, Bansal, & Chen, 2009). KPI defined measures of progress towards achieving specific performance goals, and critical to the success of the organization. A dashboard provides various information in a summary form to show how an organization is currently performing in an operational manner. In a data warehousing environment, business intelligence dashboard shows users the effectiveness of operations (Rasmussen, Bansal, & Chen, 2009). A dashboard supports a wide variety of users, from individuals to the organization as a whole. Dashboards can give a quick insight into an organization’s performance (Rasmussen, Bansal, & Chen, 2009).

Figure 2.3.2: Dashboard of Georgia State University

Note: From "Georgia State University Dashboard", http://www.gsu.edu/institutional_effectiveness/40647.html, 2010.

Production reports

These are typically large, detailed reports that have the same basic structure each time they are produced. They may be printed, or distributed online, either in Web-based reports or as formatted files (Microsoft, 2009). One advantage of a production report is that the same information can be found in the same place in each report. A production report may consist of one large report showing information about all parts of the company, or it may be "burst" into individual sections delivered to the relevant audience (Microsoft, 2009). Production reports are often used by administrators and tactical decision makers.

Figure 2.3.3: Microsoft Reporting Server

Note: From "Head Count per program", Regis University, 2010. Printed with Permission.

Analytical reports

These are dynamic, interactive reports that allow the user to "slice and dice" the information in any of thousands of ways (Microsoft, 2005). Analytical reports can display simple summations or complex calculations as with dashboard reports. They typically allow drill-down to very detailed information, or drill-up to high-level summaries (Microsoft, 2005). This type of report is typically used by analysts or "hands-on" managers who want to understand all aspects of the situation (Microsoft, 2005).

Figure 2.3.4: Analysis server

Note: From "Number of Applicants and Admitted Students", http://www.imir.iupui.edu/picx/reports/default.aspx/1/IUPUI, IUPUI-Indiana University, 2010.

Relationship of BI, data warehousing, OLAP, and Analysis Services

According to Kimball, Ralph (2008), a data warehouse is a perfect architecture for storing the data used for BI. But data in a warehouse is no use until it is transformed into the information that decision makers can use. There are several ways to convert the data in a data warehouse into information. This section explains that OLAP (online analytical processing), which is one of the best technologies for converting data into information and Analysis Services, which implements the benefits of OLAP.

OLAP (Online Analytical Processing)

In 1993, Codd introduced the term online analytical processing (OLAP) and proposed 12 criteria to define an OLAP database. Based on OLAP criteria, OLAP is a technology optimized for, complex ad-hoc queries, high level summarizing with "drill-down" through levels of detail, and interactive analysis by slicing and dicing to get different views of the data. OLAP also allows for fast data access and rapid execution of complex database queries in real time.

Figure 2.4.1: OLAP tools in the BI architecture

Note: Kimball, and Ross (2005). The Data Warehouse Toolkit: The Complete Guide to Dimensional Modeling, Second Edition. John Wiley & Sons

The one of the benefits with OLAP is a fast response consistently (Wrembel and Koncilia, 2007). Building new reports to answer questions is a slow and costly process due to multiple systems involved and in each system information is spread across multiple tables. Wrembel and Koncilia (2007) describe normal practice in work places such that there are administrators in any organization who spend a lot of hours every month or quarter exercising their Excel skills to join data manually from different systems to create simple reports. If they use OLAP tool, it will typically take less time and be more accurate. The other benefits are Metadata-based queries and Spreadsheet-style formulas (Wrembel and Koncilia, 2007). Metadata-based queries (called MDX query) allow a cross-tabulation with column headings and row headings. Metadata-based queries also allow nesting multiple layers of attributes as column headings (Wrembel and Koncilia, 2007). Spreadsheet-style formulas provide by OLAP is similar to Excel spreadsheet, but is more powerful than a simple spreadsheet formula.

The data structure that OLAP create from the relational data is called OLAP cube. OLAP cubes can be thought of as multi dimensional array. An administrator might want to analyze its enrollment data by school, by major, by geography, or something else. These different analyzing criterions are the OLAP cube dimensions.

Figure 2.4.2: OLAP Cube Data Structure

Note: Microsoft (2010), http://msdn.microsoft.com/en-us/library/aa140038(v=office.10).aspx)

Figure 2.4.3: Cube Development with Microsoft SSAS

Note: From "Financial Aid Summary" with commercially used sample data.

Once cube is developed, there are several OLAP client tools that work with the cube. Here are some examples:

Panorama NovaView – It is developed by Panorama Business Intelligence Software Company and allows displaying and navigating the levels and dimensions of a cube (Panorama, 2011).

ProClarity – It is developed by ProClarity Software Company and has an interface including visualization for browsing OLAP data (Microsoft, 2007).

Temtec Executive Viewer includes various charting and reporting options, including the use of maps to visually display OLAP data (Bloor, 2011).

Microsoft Office includes four tools that support OLAP such as Pivot Table and Pivot Chart in Excel, Pivot Table list and the Office Chart Component in the Office Web Components.

Chapter 3

Project Methodology

This chapter describes the process that was used to develop and conduct a web-based survey that would give insight into the use of data warehousing in higher education and its application to decision making. Reasons for choosing a survey approach are given, and the method for selecting respondents is explained.

Research design

Because there has been no research done on the use of data warehousing in ABC University, the purpose of the survey was mainly exploratory rather than explanatory. The survey methodology offers an ideal mechanism for gathering input from a limited number of respondents representing a cross section of departments in ABC University. The survey result will provide descriptive of state of data warehousing.

With this study, the target populations are senior decision makers, practitioners of data warehousing, information technology staffs, and institutional researcher in ABC University. The population that the survey was addressed to consists of various staffs and decision makers.

Presentation of survey

This study investigates the use of data warehouse in ABC University. The invited people for the survey are management personnel, Data analyst, administrators who are responsible to generate reports at regular bases, active users of data warehouse, and ITS/data warehousing staff regarding their perceptions of the use of data warehousing to support decision making at various management levels within the institution. The results are grouped in this analysis to correspond to my four research questions, which are current usage of DW, major drawbacks of using DW, current BI tools being used for decision making, and expectation regarding DW from ITS. The actual questions that were asked in the survey can be seen in Appendix A.

Analysis of survey data

The survey was administered from January 14 to January 30, 2011. 17 out of 28 invited participants responded, generating a 61% response rate. Based on survey results, 53% do not use the data warehouse at all. Though most would like to use the data warehouse, few have the ability to do so. 65% of respondents are eager to try new data warehouse-related software.

The concept of management level is often referred to as Anthony’s Triangle that Anthony (1965) described three levels of management – Operational, tactical, and strategic. While any operational decisions are made at the bottom, some tactual decisions in the middle and few but important strategic decisions at the top of the triangle. I targeted all three management levels in this survey.

The following list is the summary of the primary roles of participants:

Executive, dean, manager or director: 35.3%

Data Analyst or advanced user at a college unit: 23.5%

Data Analyst or advanced user at a university department: 41.2%

Table 3.1

Participation rate by management level

Note: 17 out of 28 indicated their primary role at the university. Detail in Appendix B.

Overall, 35% of respondents are managers or executives, while 65% are analysts or advanced users.

Data Warehouse Usage

Based on survey results, 53% do not use the data warehouse at all. Though most would like to use the data warehouse, few have the ability to do so. 65% of respondents are eager to try new data warehouse-related software.

Table 3.2 displays the result of question #1 about the major drawback in using data warehouse. The major drawback of the data warehouse was respondents do not know how to pull data or create reports when using the data warehouse (26%) and there is no known metadata or data dictionary for data reference (19%). The result indicates that the majority of respondents do not have proper knowledge or training on Data Warehousing.

Table 3.2

Data Warehouse Usage in ABC University

Note: Survey result #2: The major drawback in using data in the data warehouse. Detail in Appendix B.

If the respondents use the data warehouse, only 6% extensively use the data warehouse for querying and reporting purposes and 30% use the data warehouse occasionally.

0% of respondents use the data warehouse for interactive analytical processing such as dashboards, analytical cubes, or scorecards. The data warehouse is rarely used to make strategic or executive decisions, namely, to define goals or policies or determine organizational objectives.

Only 6% extensively use the data warehouse to make tactical or managerial decisions, namely, to acquire and efficiently use resources in the accomplishment of organizational goals, or establish or monitor budget. 11% use the data warehouse to make operational decisions, namely, the effective and efficient execution of specific goals using existing facilities and resources to carry out activities within budget constraints.

94% of respondents use Colleague (Transactional system) to answer questions requested by internal decision makers. Most respondents are not using the warehouse at all for specific reporting areas, such as faculty information, admissions, budgeting, and so forth. However, 35% are using the warehouse for enrollment, student records, and degree and major related reporting. Overall, very few use the warehouse to make strategic, managerial, or operational decisions (12%).

Only 24% have a ‘user sponsor’ for the data warehouse project. 88% share data with others extensively or moderately with others. Yet 59% through 88% do not use the data warehouse at all for specified purposes such as downloading or querying data.

47% state that being able to retrieve historical data is a major benefit of the data warehouse. The most common objective for using the data warehouse is to be able to report data that is frozen in time (65%). The major application tool to view the data in the data warehouse is Microsoft Excel because it is easy to use and there is no additional cost.

Since Data Warehouse is not popular system in ABC University, the concept of BI (Business Intelligence) tool is mostly unknown (88%) (See table 3.3).

Table 3.3

Popularity of BI

Note: Survey Result #20. User’s familiarity with the concept of BI. Detail in Appendix B.

Overall, respondents have realistic expectations from Information Technology Systems regarding the data warehouse, and understand they will not be getting everything they need on the first iteration of the warehouse (See table 3.4).

Table 3.4

Willingness to learn Data Warehouse

Note: Survey result #14. User’s literacy of the data warehouse. Detail in Appendix B.

Data Warehouse in Broad Subject Areas

Based on the survey, the major reasons of low adaptability with data warehousing are users’ unawareness of existence of data warehouse and its usage, and ITS personnel do not communicate with business users appropriately in order to access their needs. According to ABC University Information technology manager, Data Warehousing has been adopted by ITS in ABC University over 7 years because the IT management was quick to see the potential of data warehousing, but the business users did not have chance to be educated or trained with data warehousing. According to result of question 8 (Appendix B), the business users were not the main driver for data warehousing projects and this has led to the failure of a lot of data warehouse projects. There was often reluctance on the part of business executives to sponsor these database development projects that were sponsored by IT.

41% say that difficulty communicating with ITS regarding needs causes reluctance to use the data warehouse. 41% state other reasons in comments for reluctance, such as lack of training, data inconsistency, lack of support, slow service, or lack of data.

Before data can be stored within the warehouse, it must be cleaned, loaded, or extracted. This is a process that can take a long period of time. Users who will be working with the data warehouse must be trained to use it. If they are not trained properly, they may choose not to work within the data warehouse.

According to ABC University Information technology manger, there are only two data warehouse developers. ABC University currently does not have a person who can devote in identifying and analyzing the problem, and identifying the solutions or expectations. When a decision is made for the solution, there should be a person who designs, implements that solution, and tests the results to see if the solution solved the problem. Primarily, there is no person who can deal with business and computer-based situations. As a result, each department decided not to get a help from Information Technology Department and assigned an employee or group of employees who are generating various reports manually based on inflexible colleague system. Consequently, it is impossible to share data between departments due to the decentralization of data.

The Table 3.5 summarizes the extent of use of data warehousing across the campus for a data warehousing.

Table 3.5

Data Warehouse in Broad Subject Areas

Category/ Section: 

Data Warehouse Survey Questions/Data Warehouse Reporting Functions

Instructions:  Indicate the extent you use the data warehouse for each of the following reporting areas.

    =Extensively |     =Moderately |     =A little |     =Not at all |     =Not applicable or I do not know

Questions

Percentages

 

 

 

 

 

Faculty information or awards

-

6.3%

-

81.3%

12.5%

Alumni

-

-

6.3%

81.3%

12.5%

Advancement or development

-

-

6.3%

81.3%

12.5%

Recruiting

6.3%

-

12.5%

75%

6.3%

Admissions

6.3%

6.3%

12.5%

62.5%

12.5%

Retention or student success

-

12.5%

6.3%

68.8%

12.5%

Enrollment

12.5%

12.5%

12.5%

50%

12.5%

Financial aid

6.3%

6.3%

-

75%

12.5%

Budgeting

-

12.5%

6.3%

68.8%

12.5%

Financial / Accounting

-

6.3%

-

81.3%

12.5%

Purchasing / Accounts payable

-

-

6.3%

81.3%

12.5%

Revenue

-

6.3%

6.3%

75%

12.5%

Human resources

-

-

12.5%

81.3%

6.3%

Payroll

-

-

6.3%

87.5%

6.3%

Labor distribution

-

-

-

93.8%

6.3%

Employee retention

-

-

-

87.5%

12.5%

Student or employee demographics

6.3%

-

25%

56.3%

12.5%

Registration (students)

6.3%

-

31.3%

50%

12.5%

Student records / information

6.3%

-

31.3%

50%

12.5%

Degrees / majors

-

-

25%

62.5%

12.5%

Facilities / property / space utilization

-

-

-

87.5%

12.5%

Campus directory

-

-

6.3%

81.3%

12.5%

Faculty research or grants management

-

-

-

87.5%

12.5%

External reporting (eg. IPEDS, external surveys, etc.)

6.3%

18.8%

-

62.5%

12.5%

Internal reporting (ongoing or one time reports provided to specific groups on campus)

6.3%

18.8%

25%

43.8%

6.3%

Course or curriculum scheduling

-

-

-

86.7%

13.3%

Faculty workloads

-

-

-

87.5%

12.5%

Other area

-

6.3%

-

62.5%

31.3%

Note. Survey result #16. Scale of Extensively to Not at all, where 1=Not applicable, 2= Not at all, 3=A little, 4=Moderately, and 5=Extensively. Detail in Appendix B.

Based on performing on the data gathered from the questionnaire, the following conclusions can be highlighted in regard of the current status of data warehouse technology in the investigated university.

1. ABC University possesses mature data warehouse technology, but users mostly rely on data from transactional database (Datatel).

2. ABC University does not have any existence of premier DW project to persuade users to adopt the data warehousing technology.

3. Business Intelligence tools to make a business decision effectively have not been introduced in ABC University.

Chapter 4

Lessons Learned

This research started with question why Data Warehousing is not utilized as much as it should be in ABC University while there are a lot of benefits with a data warehousing. The research with Data Warehouse in ABC University allows an opportunity to meet various university staffs, managers, and executives who are interested in making business decisions with data warehousing. The meeting with university staffs revealed that the academic community is not just a place for students and faculties. To keep the higher education accreditation, the academic community has to produce various reports accurately and consistently for external, for strategic planning purpose, and governance on timely manner. ABC University prepares and submits numerous responses annually to a wide variety of surveys and formal requests from the public and private sectors, including federal and state agencies, as well as independent and local organizations for statistical data and information. Data warehouse technology is a great tool to defeat data-related obstacles and improve decision making initiatives. A data warehouse solution is an environment built on many software technologies and is a complex process to establish sophisticated and integrated information systems.

The results from this study revealed that top management sponsorship, existence of championship, a skillful project team, availability of resources, and end-user involvement are important considerations for ABC University to utilize the data warehousing technology. Sensitization of use of data warehouse for information gathering and decision making should be done at the various available user levels. This will encourage participation and use of data ware house by all in a better and more efficient manner (). Studies on use of data warehouse by Kafner in 2008 showed that participation and use of information from data warehouse inform of reports is positively correlated to increase in use. It is thus important to implement a data warehouse considering how all players and users of the system can best gain from the system. 6.3% of staff from ABC are able to generate reports and use the reports for decision making while a majority of the reminder mainly utilize the data warehouse at the transaction level implying they mostly enter data. To curb the above and ensure even those entering the data can be able to gain from the system it is important to understand need and information gaps existing at all user levels. This will enable customization of reports at all levels and ensure decision making is all round and relevant at all levels this as kafner showed in his study will lead to increased use and appreciation of data warehouse.

Even though a data warehouse project is more technology oriented, the goal of data warehouse is to mostly help the senior management to obtain insights into the information they already possess. Thus, top management sponsorship is very important. In typical information hierarchy models (see figure 4.1), the top level of the pyramid is usually the CEO who is looking at the information in Executive Information Systems and the person at the bottom of the pyramid is a transaction person. Hence, collecting information and trying to scope it should be done from top to bottom of the pyramid. If the high-level management supports the data warehouse projects, then needed resources such as securing required capital and human support will be obtained appropriately. To successfully ensure implementation of a data warehouse it will be necessary to have support of top management. This is only possible if top management understands benefits of the data warehouse and are able to witness how data gathered in the data ware house can be turned to information and used for sound decision making. Meta-analysis studies on impact of support from top management on implementation of new system showed that success of implementation is directly proportional to management support. Thus for a data warehouse to be successfully implemented and ensure that users at all levels appreciate it effectively it will be necessary for top management to provide sponsorship. Use of reports to inform on decision making at the top will push those at lower levels to ensure the necessary steps are available for top management to get the reports needed.

Figure 4.1: Four level pyramid models based on the different levels of hierarchy in the organization

Note: Retrieved from http://www.abitabout.com/Information+Systems

The existence of championship is considered one of the most important factors effecting the adoption of data warehouse technology (Hwang et al., 2004). A champion is people from inside the organization, who appreciate and support the adoption of new technology. They play an integral role in providing necessary information, required resources, needed assistance, political support and stimulate the staffs to cope with such technology. Pioneers play a critical role in persuading the staff to see their own personal visions to adopt new technology.

Having a skillful project team will affect the smooth development of the data warehouse project (Hwang et al., 2004). Skillful project team members should have proficiency in data warehousing matters, strong background and knowledge of it, and communication capability. It is necessary to select the members from different departments in order to add diverse values to data warehouse project as well as educate them in different aspects. This will provide necessary support for users with inadequate capacity and allow for faster growth in use as capacity is spread throughout all the relevant user department and levels.

Involving the end-users in the data warehouse project has an endless impact on promoting the vision of adopting this technology (Hwang et al., 2004). Understanding users’ needs and expectations and trying to meet them lead to reduced resistance and increased acceptance of the new technology. Appreciation of new technology is vital for realization of advancement and increased use and appreciation of system capabilities. For the aforementioned to be realized it is necessary that end users are heavily engaged in the use of the system from the early stage of set up (). If the end user identify with the system and rely on it for there daya to day work then they own the system and increase use of the sytem by others will be realized ().

There is strong connection between success factors and success rates, which leads to the conclusion that the success factors need be considered when implementing a data warehouse for the highest probability for success (Wixom et al., 2001).

Conclusion

This research revealed that changes are one of the most prominent drivers for every business including academic communities. The changes come in many different forms and from many different sources, including federal and state regulations, student interest and preparedness, faculty development, industry support, employer skill demands, industry support, employee skill demands, instruction materials and techniques and financial support. ABC University slowly realized to take an action toward the changes around them as well. One of the forms of changes in ABC University was creating a new academic unit of Institutional Research and Data Warehouse will play a significant role to support this unit.

Overall, the completion of this research provided an excellent feedback for preparing a better solution on Data Warehousing for the academic community. Data Warehousing is the best option for ABC University to provide external and internal entities with excellent, consistent, accurate, high quality, and timely information about students, alumni and the institution by faster access to more accurate and reliable data.