- Exact and Effective: ensuring that the datais reasonably accurate. We should concentrate on proving the accuracy of a dataset because we are aware that the majority of its data are valid. Even though the data is true and accurate, this does not necessarily imply that it is accurate. Finding correctness aids in determining if the information input is genuine or not. For instance, even when a customer's address is recorded in the format required, it may not necessarily be in the correct one. An extra character or value in the email renders it inaccurate or invalid. The customer's phone number is another illustration. This implies that we must rely on sources of data and double-check the information to determine its accuracy. We may be able to find a variety of resources that can assist us with cleaning, depending on the type of information we are using.
- Full Information: Completeness is the level at which we should be aware of all necessary values. Completeness is slightly harder to attain than precision or quality. We almost never have all the information we require, thus. You can enter only known facts. We can attempt to finish the data by repeating the data collection procedures, such as approaching clients again and conducting new interviews, etc. We would have to enter each customer's contact information, for instance. They might not all have email addresses though. We must leave those columns empty in this instance. We can try and enter incomplete or unknown information there if the system demands that we complete every column. However, inputting such figures does not imply that the data is correct. If the system demands that we fill up every column, we can try to insert blank or unknown words there. However, the data is not complete just because these values are entered. It would nonetheless be considered unfinished.
- Keeps Data Consistent: By contrasting two related systems, we may determine whether the data is valid between datasets or within a single dataset. To determine whether the data values inside the same dataset are consistent or not, we can additionally check them. Relationships can affect consistency. For instance, a customer's age might be 25, which would be a real and correct number, but it might also be presented in the same system as a senior person. In these circumstances, we must cross-check the data to determine which value is accurate, much like assessing accuracy. Is the client over the age of 25? Or is the customer an elderly person? A few of these figures can only be accurate. There are numerous techniques to ensure consistency in your data.
- by logging onto various systems.
- by examining the origin.
- by examining the most recent statistics.