Article

~ Possible errors

Possible errors despite care taken

The Monument’s start-up phase involved a lot of critical research on the sources and consultations with many experts. An enormous amount of data was entered into the system, and complex software was written to link all the data. Using so many different historical sources, however, increases the likelihood of error, despite the care taken. Errors may be present in the sources themselves or they may arise during the inputting or linking of data. Visitors to the website have been helping to identify and correct errors since 2005. Eliminating every single inaccuracy may be an impossible goal.

Random testing

Random testing of 100 families, carried out by an expert familiar with the material, revealed a single case (i.e. 1%) in which the family relations had been reconstructed incorrectly. The Monument contains 16,810 families consisting of three or more persons. For 1,199 of these families (i.e. 7.1%), the family relations were impossible to reconstruct without risk of error. In these cases, the family members are listed as living at the same address, but the nature of their relationship is not indicated.
Among the 330 persons in the test sample, five matching errors were identified (i.e. 1.5%). This means that the names from the register lists were linked to the wrong names or were not linked at all to names from In Memoriam. Persons may therefore be listed in the Monument under the wrong first or last name or with the wrong date of death. It is also possible that some people may not be listed in the Monument at all, due to a lack of information. The makers of the Monument are aware that these errors may cause distress, and offer their sincere apologies for them. At the same time, we would urge anyone who discovers a mistake in the Monument to inform us, so that we can correct it.

Apparent inconsistencies

There are also apparent inconsistencies that cannot be resolved. Obituaries and other documents may contain information that differs from the personal details or family composition in the left column. Both sets of data reflect the situation at a given moment in time, and they may differ. The same holds true for data in the biographical notes. These notes are based on sources that may describe a period other than the one in which the register lists were compiled. For example, the biographical notes may describe someone as an unmarried seamstress living in Deventer, while the register lists may describe the same person as a married shop assistant living in Amsterdam. Both sets of data are accurate with regard to a particular period. Discrepancies of this kind actually reveal a lot about the way the people in the Monument lived their lives, so no attempt has been made to iron out inconsistencies.

All rights reserved