Data Validation Matrix for Programming Courses

Journal of Information Systems Education, Fall 2005 by Schrage, John F

ABSTRACT

Controls are obviously needed for any application no matter how processed. Errors occur in the processing of the data with the severity dependent on the dollar relationship between the error and its results. The nature of each error, from the bit to the file usage, can be traced to specific sources. The cause frequently implies the method of choice for correcting that situation. Man, in a generic sense, has always used his thinking ability coupled with business knowledge to control and edit the manual process. These same aspects must also be used in machine processing. The computer can methodically edit and control data for applications, even to including examining character bit structures, but only in so far as human instructions are provided to the machine. To facilitate computer intervention in the data validation process, a validation matrix has been developed and used for over 25 years in this collegiate programming setting.

Keywords: computer programming, data validation

1. INTRODUCTION

The organization should be examining all data input/entry within every application so programming errors do not affect the monetary structure of the organization. Programmers should be implementing general edit techniques in all application areas. Surveying the normal procedures by which a beginning programmer is taught data validation programming and processing gives some concern. Changes to data validation appear to be happening in those organizations which emphasize interactive use of data. Some languages are adding validation key words to the language structure to aid programming; while many languages provide for data validation by the "brute force" mode of checking fields or character by character into a buffer mode. While the data validation techniques seem to be taught in many academic programs and business training settings, what is taught and what is implemented seem to differ.

2. VALIDATION MATRIX

Initially developed by a now-retired faculty member and then revised during classroom use by the author over the last 10 years, the data validation matrix has been a major topic discussed in the business programming courses and has been requested by former students that did not keep the sheet from their classes. When the author has included the data validation matrix as an aid in the application-related courses of the business area, students seems to have treated the material as "nice to know". When bad data (character, field, record, and so forth) within a programming assignment or machine related application assignments have been given, the need for better error prevention becomes obvious and the need for controls leave the "nice to know" area to become "essential needs" for the situation. Programming students have thought that data validation was "busy work" in programs and accounting students were more concerned that the application worked, but when faced with bad data, the problem of dealing with the lack of validation were evident. The programming courses have thus changed to include more validation that is tested first by the student and then by the instructor when the project appears to be finished. For our programming classes in Visual Basic, COBOL, and JAVA, the validation matrix, as shown in Table 1, is explained and demonstrated with programming examples. Some of the validation appears to be automatic based on the language and its constraints.

Each entity scope in the table, with the nature of the error condition, is described in the language specific course with an example. For example, when a numeric field is specified/defined in the graphic interface of Visual Basic or COBOL and then a non-numeric character is attempted to be entered, the field rejects the contents [based on the context nature field notation from the validation table] and expects the data entry person to enter the correct response - a number. No message is given just a beep and the cursor is sent back to that text box. The TRY -- CATCH aspect in Visual Basic heavily uses the context and content field error conditions to validate data entered in the user interface portion of the programming code. Other languages and settings need code to be written by that programmer such that the data processing does not attempt to accept the entry.

3. CONCLUSION

Data validation has to become a common aspect in the training of beginning business people, especially those planning to obtain a position in the computer programming area. The need for editing in the microcomputer environment is probably more important than mainframe in that the backing up of files is not as prevalent as the larger installation. Educational programs should be stressing the implementation of controls within applications, especially in the design and programming courses. The data validation matrix chart has been a key point in programming courses for over 25 years in our business oriented degree program. The chart has been updated several times since first being used and also published in an accounting journal.