THE DEVELOPMENT OF A SYSTEM TO TRACK CONSTRUCTION GRADUATES

INTRODUCTION

During the summer of 1987, the Department of Building Science at Auburn University set in motion a plan to develop a computerized database. The primary purpose was to provide departmental access to detailed current information on graduates from the Building Science curriculum. A secondary purpose was to provide student access to some of the more general information that would be compiled.

For several years the department head had been interested in mailing out a questionnaire with the intent of gaining some information regarding graduates of the program. Questionnaires of this type can provide insight into several areas that must be evaluated, especially when preparing documentation for accreditation visits.

At an early stage, it was decided to expand the scope of the project from a standard ques­tionnaire to a computerized database. Even though this would require substantially more work, we felt that we would eventually have to set up a computerized system. This seemed to be the time to do it.

STUDENT INVOLVEMENT

The initial step was to ask the members of the student honorary fraternity, Sigma Lambda Chi, if they would assume a portion of the job as a project. The students were enthusiastic and agreed to help. Interestingly, they were already in the process of compiling a departmental newsletter that was to be mailed at approximately the same time that we hoped to send the questionnaire.

The students organized themselves into two groups, one group concentrating on the news­letter, the other on the database. A faculty member would develop the questionnaire based on input from the department head.

At this time it was decided to mail the news­letter with the questionnaire. We felt that this would serve two purposes. First, it would help to defray mailing expenses and secondly, the fact that an alumnus received a newsletter might make him more receptive to returning the questionnaire.

The first task was to obtain a list of names and addresses of graduates. For a small fee, the alumni office furnished a list of 1350 (out of approximately 2000) graduates. Unfortunately, but not unexpectedly, many of the addresses were no longer current or were simply addresses of parents. Since we could not develop a more accurate listing, questionnaires were mailed to all 1350 people on the list.

DEVELOPMENT OF QUESTIONNAIRE

While quite simple in appearance, the ques­tionnaire (Fi ure 1) and the accompanying cover letter Figure 2) required a good deal of forethought and planning. Before drafting these documents, we looked at several other questionnaires (some given to us by other ASC schools). The primary goal was to maxi­mize the number of responses.

Since some of the questions were related to salaries and other somewhat confidential information, we felt a cover letter offering some explanation was necessary.

In the hopes of eliciting a larger response, we decided to limit the questionnaire to one page in length, even though we were forced to omit some questions. We devised a system that would allow the person to supply a numerical response whenever possible. This would make compilation and tabulation of the data much easier.

DEVELOPMENT OF DATABASE

As the questionnaire and newsletter were being developed, a team of students was busy developing the computerized database. We chose dBase3+ for the software because several of the students were familiar with it, and because it was available in our computer lab.

MAILING AND CATALOGUING

The questionnaires were sent out one month in advance of the requested return date. We considered using bulk mail to reduce our costs but were concerned about two and three week delivery times. We purchased printed mailing labels from the alumni office, and the students placed them on the envelopes. At the same time, another group of students placed a letter, questionnaire, and newsletter into each of the 1300 envelopes. This took more time than anticipated.

The first responses began to appear within one week. As the responses arrived, they were stamped with a date and filed in a binder. Approximately 300 of the responses were returned by the requested deadline. The last response was received over 4 months after this date. A total of 386 responses were returned.

DATA TRANSFER AND COMPILATION

After most of the responses had arrived, they were keyed into the computers by the Sigma Lambda Chi members. These students developed a customized input format that enabled persons who were unfamiliar with personal computers to easily transfer information from the handwritten responses to the computer. What we thought would be a relatively simple task still required some careful planning. We reserved our PC lab containing 14 IBM PCs for two 2-hour periods. Twenty-two students were present. The database "program" had previously been installed on each of the 14 machines by one of the students. Explanations were given to the entire group, and two students circulated the entire time answering questions. Even with more than twenty people working, it took four hours to transfer all the responses to

the computer. Depending on the skill of the typist, 6-10 minutes were required for each response.

After the responses were typed into the computer, the database had to be consolidated from fourteen disks to one. This operation is simple but does require some rudimentary

knowledge of dBase3+.         We were surprised to find that the data from all 386 responses

fit very easily on one 360 K disk. Backup copies were then made and stored in secure

places. Once all the data was placed on a single disk, it is a relatively simple task to sort, index, count, or average any of the various parameters.

The final step in the entire procedure was to obtain printed copies of the data. Printouts were made with the data arranged in every conceivable manner. We made listings by name, state, employer, and year graduated. These printouts were neatly organized and filed

in the departmental office. A second copy (with confidential data deleted) is kept on file in the library for student use.

COST

The cost of developing a database of this type is significant. It would have been very difficult without the help of our Sigma Lambda Chi chapter. A detailed cost breakdown is listed below. Note that the total cost was nearly one dollar per questionnaire.

Printing:

ERRORS

In spite of all our planning, we still made two errors in the design of the questionnaire. First, we should have included designations for military and government on question #4. This caused some problems when we were attempting to determine average starting salaries, because the entry level military salary is somewhat lower than the civilian. Secondly, we should have allowed the alumni to respond with a number (instead of a name) for the first part of question #7.

POTENTIAL BENEFITS OF DATABASE

There are many advantages of possessing a computerized listing of alumni. A few of the obvious ones are listed below.

A SUMMARY OF FACTS GLEANED FROM RESPONSES

CONCLUSION

The computerized database that was created is certainly of significant value to the department. The development of the database was more arduous in terms of manpower and more expensive than we originally expected. The involvement of a student organization in the project is almost a necessity when large numbers of responses are collected. The establishment of this type of database should not be a onetime effort but be updated on a yearly basis. The information obtained was well worth the cost and effort.

ACKNOWLEDGEMENT

The author wishes to thank the members of the Sigma Lambda Chi honorary fraternity for their good humored help in assembling the data, and Mrs. Jane Parrotte for invaluable competence in the typing and editing of this document.