The ECN No Name Newsletter is no longer being published. This is an archived issue.
[previous article] [next article]Graduate students and faculty members are often involved in a research project that requires the use of statistics. Researchers may need help designing their experiment or may require some assistance in analyzing or interpreting results. They may simply need to perform some simple descriptive statistics or to create a graph which displays their findings in an understandable form. There are two services to which they can turn--the Statistical Consulting Service and/or SAS software.
The Purdue University Statistical Consulting Service is supported by the Statistics Department and PUCC and is supervised by George McCabe (MATH 536). It provides the University community with free advice concerning problems involving the use of statistics. Individuals needing assistance with statistical computer software or seeking help with simple statistical problems can drop in at the Statistical Consulting Service during normal working hours. Those with a problem involving the design of experiments, the statistical analysis of data and the interpretation of results need to make an appointment with the Statistics Department.
In 1989 the service received 20 requests from faculty, 65 from graduate students and 15 from other University staff. About 46 percent of these requests came from either the School of Liberal Arts or the School of Consumer and Family Sciences. Requests were also received from the Schools of Engineering. For example, earlier this year a researcher from the School of Industrial Engineering needed to compare a new decision making methodology for assigning tasks with a standard methodology. The Statistical Consulting Service was able to run tests to determine if any differences existed between the two methodologies.
Anyone who wishes to request consulting help should contact Teena Seele in the Math Building (room 537, phone 45324) during normal working hours to receive a form. When the completed form is received, the project will be assigned to a graduate student. This process usually takes only a couple of weeks. However the last few weeks of the semester are the busiest for the service and persons who turn in their forms during this time may have to wait until the following semester for their projects to be assigned.
Once the project is assigned to a student, an initial meeting will be arranged. Any graduate student using the service is required to bring his professor or research adviser for this initial meeting.
SAS software is another method that students and faculty can use to solve their statistical problems. SAS has been used at Purdue for about five years. This semester SAS 6.03 is being placed online at the ECN.
What is SAS software? The SAS system is an integrated system of software products that can perform an amazing amount of tasks including:
The SAS software can both read and write data in virtually any form. It can perform practically any statistical or mathematical operation and can also create bar charts, pie charts, plots, and maps of every country in the world and every state in the country. One student in the statistical consulting office told me that SAS can do his laundry for him. That might be pushing it, but, generally, if you want to do anything mathematical or statistical, SAS can get the job done.
SAS is not just software for statisticians. It is used as a tool to analyze data. Those who use SAS may not be statisticians themselves. The graduate student writing his thesis can use SAS to analyze and interpret his experimental data. SAS will enable him to display this data in many different graphical forms as well.
Both the Statistical Consulting Service and the SAS software can be an invaluable aid to the student or faculty member with problems involving the use of statistics. Manuals explaining the various SAS programs are borrowable from Potter Library reference desk. Short courses assisting new SAS user's to use the statistical software are available. Type "help ecnnews" and pick ShortCourses.
HTML