Originally posted by JeffHamm
View Post
https://www.frontiersin.org/articles...019.00703/full
Ultimately the study suggests that you can assess the accuracy of testimonies through understanding the speed and confidence of how the witness answers questions. If audio recordings of witness testimonies were available then, you could use this method to get a sense of which details are more likely to be true versus the more spurious or false memory details.
The accuracy in the study was quite impressive. I would not be surprised if these techniques make their way into law enforcement.
For our purposes, a matrix of some kind where we grade the details provided by witnesses based on the wording would be the next thing. "I think, could have been, probably, maybe" we weight low. "Definitely, absolutely, certain," we grade with a higher weighting. Details that do not seem either confident or unsure are weighted mid level.
Perhaps eliminating newspaper (with exception of police newspapers) testimonies from the outset for the reasons you outlined as well, focus on inquest testimonies and police reports / letters. It would be interesting to also include a column that showed if that detail is corroborated by another witness. That would give more weighting to that detail.
From that you grade details with a certain accuracy based on confidence and corroboration and the results could be:
High Weighting:
Suspect Moustache Colour
Suspect Height
Position of Victim's body
Mid Weighting:
Time of Body Discovery
Low Weighting:
Suspect Hair colour
Last Person Seen With Victim
Hat of Suspect
Clothing of Suspect
etc
Just a thought.
Leave a comment: