Errors and Forensic Science
Errors in forensic science are not new. We can all cite miscarriages of justice in which forensic science has been involved to a greater or lesser extent. Error rates, validity and reliability are subject to much current discussion. But is forensic science particularly error prone? How does it stack up with other fields of endeavor? What do we actually mean when we speak of ‘error’? And what can be done to minimize error?
The word ‘error’ has a number of meanings which can be broadly classified as:
- Measurement error. This term is better described as measurement variation as error implies a mistake and this form of error is not a mistake. Measurement error is the difference between a measured quantity and its true value. Every time a measurement is repeated there will be a measurable variation in the result generated. The size of these variations depend in part in the sensitivity and precision of the form of measurement involved. The degree of variation for a measurement can be determined during validation studies and acceptable values determined for each form of measurement.
- Inadvertent errors, which are mistakes, in that they are due to a deviation from the expected. These errors can be sporadic one-off events or systematic (for example errors caused by a mis-calibrated instrument that affects all measurements or from an operator misunderstanding an instruction and routinely performing the operation incorrectly).
- Type one error or false positive. For example, concluding that a substance is present when it is not present.
- Type two error or false negative. For example, concluding that a substance is not present when it is present.
- Deliberate errors or dishonestly.
Note: an observed result may involve one or more of these errors – they are not mutually exclusive.
The descriptions above are not restricted to forensic science, they are common to all fields of human endeavor. For example:
- a 2016 report of the Victorian Auditor-General’s Office discusses concerns regarding ‘Patient Safety in Victorian Public Hospitals’;
- the 2015/16 Annual Report of the Victorian Director of Public Prosecutions discusses their involvement in appeals lodged with respect to both sentence and conviction. The outcomes of the appellate process included finding where the original decision was overturned or amended.
So, if error to common to all fields of human endeavor, what can be done about it? These are my suggestions:
- understand and recognize the difference between a measurement error and a mistake;
- understand and recognize that mistakes occur in all fields of human endeavor;
- understand and recognize that mistakes cannot be totally eliminated in any system;
- understand and recognize how mistakes can be minimized and implement appropriate systems;
- understand and recognize how to detect error and how to mitigate or rectify the consequences; and
- use error detection as a tool for further improvement of the system.
Many of you will have recognized many aspect of quality management in the above list. This is not a coincidence!
If you are interested in further information about Error Management and Forensic Science, the National Institute of Standards and Technology (NIST) has conducted a number of International Symposia on Forensic Science Error Management. Details of the 2017 Symposium including all the presentations can be accessed here.
Remember: To err is human; to forgive, divine. Alexander Pope (1688 – 1744).