Papers presented in SAS Forum

2007

1) Panel Data Analysis Using SAS

This paper demonstrates the panel data modeling facilities that can be found in SAS/ETS PROC TSCSREG and SAS Enterprise Guide (EG). The procedure in TSCSREG is very useful and powerful for econometric modeling especially when cross sectional and time series data are combined. However, for novice user the SAS Enterprise Guide is highly recommended since one does not need to use SAS codes. In this paper data set from the manufacturing sector is used and the level of productivity is analyzed. The results show that both techniques produce good and consistent estimation results when compared to other software for panel data analysis. However, the advantages of using the SAS Enterprise Guide (EG) are: it is menu driven (point and click), the codes are reusable and the outputs are very nice.

Full Paper

2) Activity-Based Management in Higher Education: Can it Work?

The financial management of a university is becoming increasingly complex. Recent cuts in government funding, greater student enrolments, expectations for better outcomes from their education and for programs that are tailored to the market needs, are demanding public universities to become more cost effective. University managements thus need to scrutinize their spending and demonstrate that resources are being used effectively and efficiently. Universities must strive to work on controlling costs while aggressively working to address faculty and administrative productivity. To achieve this, university needs a new costing methodology such as activity-based costing (ABC) that can link the accounting data to the university strategic plan as well as performance measures for better resource allocation. Universities can use the ABC method to identify how faculty members spend their time by activity area and the funds allocated by activity. Despite the increasing interest in ABC in recent years, very little have been reported regarding ABC use in institutions of higher learning. Based on a case study, this paper provides a discussion of how ABC approach can be implemented in the specific context of Malaysian university.

Full Paper

3) Academic Business Intelligence System Development Using SAS Tools

Managing an organization requires access to information in order to monitor the activities and to assess the performance. In addition, information demand, data volumes, and customer populations are growing and will continue to grow exponentially. Business Intelligence (BI) solutions provide organizations with timely, integrated information that is crucial to the understanding the business environment and customer needs. These BI systems allow an organization to gather, store, access and analyze corporate data sources for business planning and decision-making. In academic institutions, the management demands more information for academic resources planning and academic excellence. This study focuses in defining and developing data warehouse (DW) for academic domain in Universiti Utara Malaysia (UUM). The dimensional model (DM) of DW in student subject area has been defined before the DW was created. A prototype of BI application based on the proposed DW model was developed and linked to the UUM information portal for accessibility. In particular, the proposed model becomes a guideline for the practitioners to develop BI system for academic domain and BI system in general. SAS ETL and SAS Enterprise Guide have been used successfully in developing the prototype of BI system. Both approaches and techniques in data gathering, transformation, loading and analyzing by using SAS tools will be focused and explained.

Full Paper

4) Construction of a User Profiling System for SAS Mobile OLAP

User profiling is the act of building up a profile of the users to have an accurate idea of what they want to do. In this research, we apply the user profile creation into an OLAP context under a mobile environment with the ultimate goal of constructing a User Profiling System (UPS). A UPS personalizes the query sent to the server according to the user’s preferences. In this paper, we introduce the SAS Mobile OLAP (SMO) project. First, we present several applications of user profiling, the methods used to construct a profile and an evaluation method for such systems. Then, the results obtained are projected on an OLAP query’s customization system. We finally attempt to propose an implementation using the SAS platform in a mobile environment.

Full Paper

5) Predicting Students' Academic Performance: Comparing Artificial Neural Network, Decision Tree and Linear Regression

Predicting students' academic performance is critical for educational institutions because strategic programs can be planned in improving or maintaining students’ performance during their period of studies in the institutions. The performance of the academic performance in this study is measured by their cumulative grade point average (CGPA) upon graduating. In this study, the students’ demographic profile and the CGPA for the first semester of the undergraduate studies are used as the predictor variable for the students’ academic performance in the under-graduate degree program. Three predictive models have been developed using SAS Enterprise Miner, that are, artificial neural network, decision tree and linear regression. The result of this study shows that all of the three models produce more than 80% accuracy. It also shows that artificial neural network outperforms the other two models.

Full Paper