One interesting point was on the notion of memory conformity when witness memory becomes tainted by discussions with other witnesses or becomes familiar with other details of the case. This reminded me of the findings of Solomon Asch and the experiment whereby participants were instructed to find the 'longest' out of three lines. All of the group but one were instructed to be accurate in the first few instances and then go wrong. The actual participant of the study went along with the group and ignored his own sense much of the time.
Conforming in this way could be very dangerous for law and thus forensic psychologists can aid in eliminating this.
Some surprising stats from the article:
"Miscarriages are not as rare as some may think. Using the conservative and internationally accepted estimate that 1% of all criminal convictions are incorrect, there could be up to 330 people wrongfully convicted in Australia each year for serious (district/county or supreme court) offences.
This figure rises to more than 8,500 if all court appearances (including magistrates’ court) are considered."