The accuracy of data entered into a computer is of utmost importance in data entry. Manual data entry has the potential to be more accurate than automated data entry. A human worker can recognise and correct errors as they occur and so an attentive data entry worker may make fewer errors than an automated system. Manual data entry does, however, introduce the possibility of human error; workers who do not pay close attention to detail or who are suffering from a lack of concentration or fatigue may make more errors.
Despite high automation, much data is typed into the databases by people through various forms and interfaces. The most common source of data inaccuracy is that the person manually entering the data just makes a mistake. People mistype; they choose the wrong entry from the list or enter the right data value into the wrong box. When capturing data, think carefully about its final destination and what rules and processes you need to put in place to upload efficiently and with high quality.
Accurate information is able to give the required results and the output necessary in either decision making giving complete and reliable information that is comprehensible for analysis.
Characteristics of Good Quality Information (ACCURATE).
Characteristics of good quality information can be defined as an acronym ACCURATE.
These characteristics are interrelated; focusing on one automatically leads to focusing on another:
- Accurate
- Information should be fair and free from bias. It should not have any arithmetical or grammatical errors. Information that comes directly or in written form is likely to be more reliable than it comes from indirectly (from hands to hands) or verbally which can be later retracted. To make sure your information is accurate; you have to make sure that all the vital information fields are completed when you enter data onto the system.
- Complete
- The accuracy of information is just not enough. It should also be complete which means facts and figures should not be missing or concealed. Telling the truth but not wholly is of no use.
- Cost-beneficial
- Information should be analysed for its benefits against the cost of obtaining it. In the business context, it is not worthwhile to spend money on information that even cannot recover its costs leading to loss each time that information is obtained. In other contexts, such as hospitals it would be useful to get information even if it has no financial benefits due to the nature of the business and expectations of society from it.
- User-targeted
- Information should be communicated in the style, format, detail, and complexity that address the needs of users of the information. For example, senior managers need brief reports that enable them to understand the position and performance of the business at a glance, while operational managers need detailed information that enables them to make day-to-day decisions.
- Relevant
- Information should be communicated to the right person. It means a person who has some control over decisions is expected to come out from obtaining the information.
- Authoritative
- Information should come from reliable sources. It depends on the qualifications experience and past performance of the person communicating the information.
- Timely
- Information should be communicated in time so that the receiver of the information has enough time to decide appropriate actions based on the information received. The information that communicates details of past events earlier in time is of less importance than recently issued information like newspapers. What is timely information depends from situation to situation. Selection of appropriate channels of communication is a key skill to achieve.
- Easy to Use
- Information should be understandable to the users. Style, sentence structure, and jargon should be used keeping the receiver in mind. If the report is targeted to newcomers in the field, then it should explain the technical jargon used in the report.
To help data capture complete vital fields correctly; the designer can add rules to fields when they create the database. Rules could include making fields compulsory to complete. The data capture will be unable to move on to the next field until this field is completed. Formatting rules require fields to be formatted as either alphabetical numerical or even alphanumerical. A date field could force the capturer to use the correct date sequence and even decide if it should be a long date or a short date format. Fields can also have a required length. This is very useful when you need data captures to complete fields like ID numbers or account numbers.
By completing all vital fields accurately the company will be able to use the information in the database to do focused marketing, sales promotions, and customer-specific offerings.
Refer to incomplete or unauthorised source documents for clarification
A common data entry problem is handling missing values. Users may assign the same blank value to various types of missing values. When “blank” is not allowed, users often enter meaningless value substitutes. Default values in data entry forms are often left untouched. The first entry in any list box is selected more often than any other entry.
Good data entry forms and instructions somewhat mitigate data entry problems. In an ideal fantasy world, data entry is as easy for the user as possible: fields are labelled and organised clearly, data entry repetitions are eliminated, and data is not required when it is not yet available or is already forgotten. The reality of data entry, however, is not that rosy. Manual data entry will always remain a significant cause of data problems.
If the source document is “incomplete”, there are some fields that are required and must be filled in before the document may be processed any further. It is important that a process is in place to ensure that the data capturer will refer to the incomplete source documents, rather than enter bogus information to complete the form.
Garbage in garbage out. Incomplete or incorrect information in a database renders the database useless. Many companies run regular data quality checks to ensure that the data is correct and all information is included in the database.
Prevention is still better than cure. A referral is a procedure used to resolve issues such as incomplete source documents or unauthorised source documents. The referral could ask the sales representative to follow up with the customer to complete the required information. If some of the source documents are unauthorised, this should be approved by management as acceptable.
Both these processes should be documented and available to all data capturers to ensure that they deal with these problems consistently.