1. Eyewitness memory can become contaminated by viewing images of a suspect before the suspect appears in a police lineup. An eyewitness's memory is least likely to be contaminated on a first test administered following best-practice guidelines soon after the witnessed event and before the witness could be exposed to biasing information

​

2. Before putting a suspect in a police lineup, there should be independent evidence of guilt.

​

3. A suspect should not be put into an eyewitness lineup procedure based solely on their similarity to a surveillance image of the culprit (or a composite sketch of the culprit). Finding a suspect in a Facial Recognition Technology search risks leading to investigation of an innocent suspect who resembles the true culprit and increases the risk of misidentification.

​

4. Under proper testing conditions, suspect IDs made with high confidence are more likely to be accurate than suspect IDs made with low confidence.

​

5. The risk of contamination is lowest on the first identification test. There are no second chances to test a witness’s uncontaminated memory of the suspect because the first test itself contaminates the witness’s memory by exposing them to the suspect in the lineup.

​

6. Nonidentifications of the suspect are probative of innocence.