For the sake of example, we will change the PHONENO field from Text to Integer. Since this type of alteration is not supported using ALTER TABLE, a table rebuild is required. See the Troubleshooting section to avoid numeric overflows and invalid date errors. The upper row shows current Access field properties, while the lower row shows matching DB2 column properties. The existing DB2 column properties will be replaced with the corresponding Access property.
| Setting | Effect on Column Data |
| Data Safe | All existing data will be retained after the operation completes. |
| Will Lose Data | The data type change will result in a loss of all column data. |
| Possible Invalid Data | Errors may be encountered during data type conversion. For example, if a CHAR column is changed to DATE and some existing data in the CHAR column does not represent a valid date, then if the extended UDF conversion functions are activated the column is set to NULL, otherwise the operation is aborted. |
| May Lose Data | Some data may be lost when data in the source column exceeds the size of the target column. For example, CHAR(50) is changed to CHAR(40). If the target column is numeric and the extended conversion functions are activated, then data is retained when the target column is large enough to contain the entire number, otherwise the target column is set to NULL. |
| Possible Overflow | If the extended conversion functions are NOT activated and the target column is not large enough contain the number, the operation will be aborted. |