There are a few kinds of standard errors that are often experienced when doing data entry. Two of the most common of these are transcription errors and transposition errors.
Every time data is manually entered into the system, there’s a possibility that an error will be committed.
Human errors do occur, and there are a variety of different reasons as to why they happen. One could be that the person misunderstood what was written or what was said. Another reason for an error was that the person was rushing, and didn’t pay enough attention to detail.
Long codes that have no significant meaning to the encoder are susceptible to error.
An example of a transcription error could be somebody entering ‘fate’ instead of ‘faith’.
Transposition errors happen when the encoder accidentally mixed up the order of numbers or letters.
For example, 78 might be entered as 87, or ‘faith’ might be entered as ‘faiht’.
Validation procedures cannot make sure that the data entered is correct—it can only check that it is rational, logical, and acceptable. It is obviously ideal to have as much accurate information as possible in your database.
Verification can be done to ensure that the data in the database has as few errors as possible. Another way to phrase this is by saying that verification is done to make sure that the data entered is equal to the data from the original source.
Verification means to check that the data from the original source document is exactly the same as the data that you have entered into the system.
Methods of Verification
Double entry – This refers to inputting the data twice and comparing the two entries.
- A classic example would be when creating a new password. You are often asked to enter the password twice. This lets the computer verify that data entry is exactly the same for both instances, and that no error has been committed. The first entry is verified against the second entry by matching them.
- While this may be useful in identifying many errors, it is not practical for large amounts of data. Here are some disadvantages of double entry:
- It would take an encoder a lot of time to input the data twice. It doubles the workload as well as the cost.
- An encoder could input the same error twice, and it wouldn’t be noted as an error.
- The possibility of having two (correct) versions of the same data exists, and double entry can’t account for this eventuality.
Proofreading data – This process requires another person checking the data entry against the original document. This is tedious as well as costly.
Checking the data on the screen against the original paper document – This can help identify transcription and transposition errors. It also saves time, in comparison with the double entry technique. However, it is difficult to keep shifting your eyes back and forth from the monitor to the hard copy, and this difficulty can exacerbate human factors such as tiredness and blurry eyes, resulting in missed errors.
Printing out a copy of the data and comparing the printout to the original paper document – This is probably the simplest method of verification because you can put both copies side by side and scan both for errors. However, it can be tedious if there is a large amount of data to check. Also, if scanned too quickly, errors could well be overlooked.
Getting a helping hand – If you verify data with a team member, then one good way of identifying errors is for the coworker to read the input data while you check it against the original document. This can consume a lot of time and it utilises two people, so whether or not it’s a good solution for any given situation depends on how crucial the data is versus time spent, and other resources (e.g. money). It is recommended that the second person reads the input data instead of yourself, because there is a high possibility that you will commit the same mistake twice; for example while transposing a number.