In this section, we apply the control framework to the generic system described above. Figure 9.4 presents a completed control matrix for the systems flowchart shown in Figure 9.3. Through the symbols P-1, P-2, . . . P-7, we have annotated the flowchart to show where specific control plans are implemented. We also have one control plan that we assume is missing (code M-1) because the narrative did not mention it specifically. The UC and UA columns in the matrix have been shaded to emphasize that they do not apply to this analysis because there is no update of any master data in Figure 9.3.
As you recall from the previous section, step 2 in preparing a control matrix is to tailor the control goals across the top of the matrix to the particular business process under review. Because our model system does not show a specific system such as cash receipts, inventory, or the like, we cannot really perform the tailoring step. Therefore, in Figure 9.4 under the operations process section, we have shown only one operations process goal for illustrative purposes. We identify that goal as goal A: To ensure timely processing of (blank) event data (whatever those data happen to be). In the business process chapters (THE “ORDER-TO-CASH” PROCESS: PART I, MARKETING AND SALES (M/S) through THE BUSINESS REPORTING (BR) PROCESS) you will see how to tailor the goals to the systems discussed in those chapters.
The recommended control plans listed in the first column in Figure 9.4 are representative of those commonly associated with controlling the data entry process. The purpose of this presentation is to give you a sense of the multitude of control plans available for controlling such systems. The plans are not unique to a specific system such as sales, billing, cash receipts, and so forth. Rather, they apply to any data entry process. Therefore, when the technology of a system is appropriate, these controls are incorporated into the list of recommended control plans (step 3 in “Steps in Preparing
Let’s take a general look at how several of the control plans work.1 Then, in Table 9.2, we explain each of the cell entries in the control matrix. As you study the control plans, be sure to see where they are located on the systems flowchart.
P-1: Document design. Document design is a control plan in which a source document is designed in such as way to make it easier to prepare initially and to input data from later. We designate this as a present plan because we assume that the organization has designed this document to facilitate the data preparation and data entry processes.
P-2: Written approvals. A written approval takes the form of a signature or initials on a document to indicate that a person has authorized the event.
P-3: Preformatted screens. Preformatted screens control the entry of data by defining the acceptable format of each data field. For example, the screen might require users to key in exactly nine alphabetic characters in one field and exactly five numerals in another field. To facilitate the data entry process, the cursor may automatically move to the next field on the screen. And the program may require that certain fields be completed, thus preventing the user from omitting any mandatorydata sets. Finally, the system may automatically populate certain fields with data, such as the current date and default shipping methods, sales tax rates, and other terms of a business event. Automatic population reduces the number of keystrokes required, making data entry quicker and more efficient. With fewer keystrokes and by utilizing the default data, fewer keying mistakes are expected. To ensure that the system has not provided inappropriate defaults, the clerk must compare the data provided by the system with that on the input.
P-4: Online prompting. Online prompting asks the user for input or asks questions that the user must answer. For example, after entering all the input data for a particular customer sales order, you might be presented three options: (A)ccept the order, (E)dit the order, or (R)eject the order. By requiring you to stop and “accept” the order, online prompting is, in a sense, advising you to check your data entries before moving on. Many systems provide context-sensitive help whereby the user is automatically provided with, or can ask for, descriptions of data to be entered into each input field. Another way to provide choices for a field or to limit allowable choices is to restrict entry to the contents of a list that pops up. For example, a list of state abbreviations can be provided from which the user chooses the appropriate two-letter abbreviation.
P-5: Programmed edit checks. Programmed edit checks are edits automatically performed by data entry programs upon entry of the input data. Erroneous data may be highlighted on the terminal screen to allow the operator to take corrective action immediately. Programmed edits can highlight actual or potential input errors, and allows them to be corrected quickly and efficiently. The most common types of programmed edit checks are the following:
- Reasonableness checks.Reasonableness checks, also known as limit checks,test whether the contents (e.g., values) of the data entered fall within predetermined limits. The limits may describe a standard range (e.g., customer numbers must be between 0001 and 5000, months must be 01 to 12), or maximum values (e.g., no normal hours worked greater than 40 and no overtime hours greater than 20).
- Document/record hash totals. Document/record hash totals are a summary of any numeric data field within the input document or record, such as item numbers or quantities on a customer order. The totaling of these numbers typically serves no purpose other than as a control. Calculated before and then again after entry of the document or record, this total can be used to determine that the applicable fields were all entered and were entered correctly.
- Mathematical accuracy checks.This edit compares calculations performed manually to those performed by the computer to determine if a document has been entered correctly. For this check, the user might enter the individual items (e.g., quantity purchased, unit cost, tax, shipping cost) on a document, such as an invoice, and the total for that document. Then, the computer adds up the individual items and compares that total to the one input by the user. If they don’t agree, something has likely been entered erroneously. Alternatively, the user can review the computer calculations and compare them to totals prepared before input.
- Check digit verification. In many processes, an extra digit—a check digit—is included in the identification number of entities such as customers and vendors. More than likely you have a check digit as part of the ID on your ATM card. The check digit is calculated originally by applying a complicated and secret formula to an identification number; the check digit then is appended to the identification number. For instance, the digit 6 might be appended to the customer code 123 so that the entire ID becomes 1236. In this highly oversimplified example, the digit 6 was derived by adding together the digits 1, 2, and 3. Whenever the identification number is entered later by a data entry person, the computer program applies the mathematical formula to verify the check digit. In our illustration, if the ID were input as 1246, the entry would be rejected because the digits 1, 2, and 4 do not add up to 6. You are already saying to yourself, “But what about a transposition like 1326?” This entry would be accepted because of the simple method of calculating the check digit. In practice, check digits are assigned by using much more sophisticated formulas than simple cross-addition; those formulas are designed to detect a variety of input errors, including transpositions.
What are two common programmed edit checks? Describe each check.
P-6: Interactive feedback checks. An interactive feedback check is a control in which the data entry program informs the user that the input has been accepted and recorded. The program may flash a message on the screen telling a user that the input has been accepted for processing.
M-1: Key verification. With key verification documents are typed by one individual and retyped by a second individual. The data entry software compares the second entry to the first entry. If there are differences, it is assumed that one person misread or mistyped the data. Someone, perhaps a supervisor or the second clerk, would determine which typing was correct, the first or the second, and make corrections as appropriate.
P-7: Procedures for rejected inputs. Procedures for rejected inputs are designed to ensure that erroneous data—not accepted for processing—are corrected and resubmitted for processing. To make sure that the corrected input does not still contain errors, the corrected input data should undergo all routines through which the input was processed originally. A “suspense file” of rejected inputs is often retained (manually or by the computer) to ensure timely clearing of rejected items. To reduce the clutter in the simple flowcharts in this text, we often depict such routines with an annotation “Error routine not shown.”
P-1: Document design.
Operations process goal A, Efficient employment of resources: A well-designed document can be completed more quickly and can be prepared with less effort.
Input accuracy: We tend to fill in a well-designed document completely and legibly. And, if a document is legible, data entry errors occur less frequently.
P-2: Written approvals.
Input validity: By checking to see that approvals are present on all input documents, we reduce the possibility that invalid (unauthorized) event data will be input.
P-3: Preformatted screens.
Operations process goal A, Efficient employment of resources: By structuring the data entry process, automatically populating fields, and by preventing errors, preformatted screens simplify data input and save time, allowing a user to input more data over a period of time.
Input accuracy: As each data field is completed on a preformatted screen, the cursor moves to the next field on the screen, thus preventing the user from omitting any required data set. The data for fields that are automatically populated need not be manually entered, thus reducing input errors. Incorrectly formatted fields are rejected.
P-4: Online prompting.
Operations process goal A, Efficient employment of resources: By asking questions and providing online guidance, this plan ensures a quicker data entry process and allows the user to input more data over a period of time.
Input accuracy: The online guidance should reduce input errors.
P-5: Programmed edit checks.
Operations process goal A, Efficient employment of resources: Event data can be processed on a more timely basis and at a lower cost if errors are detected and prevented from entering the system in the first place.
Input accuracy: The edits identify erroneous or suspect data and reduce input errors.
P-6: Interactive feedback checks.
Input completeness: By advising the user that input has been accepted, interactive feedback checks help ensure input completeness.
M-1: Key verification.
Input accuracy: By having one data entry person type the data and a second person retype that same data, we should detect the majority of keying errors.
P-7: Procedures for rejected inputs.
Input completeness: The rejection procedures (i.e., “Error routine not shown” annotations in Figure 9.3) are designed to ensure that erroneous data—not accepted for processing—are corrected and resubmitted for processing.
Explanation of control matrix cell entries.Armed with an understanding of the mechanics of certain control plans, let’s now turn our attention to Table 9.2— Explanation of Cell Entries for Control Matrix in Figure 9.4. See whether you agree with (and understand) the relationship between each plan and the goal(s) that it addresses. Remember that your ability to explain the relationships between plans and goals is more important than your memorization of the cell entries themselves.
How does each control plan listed in the control matrix in Figure 9.4work?