Crime topic 2 assessment Flashcards
What is sampling bias?
When a sample in a study doesn’t accurately reflect the target population because certain characteristics were excluded from the sample. So the results may not be able to be generalised to whole population
Evaluate research into processing/ collection of forensic evidence in terms of sampling bias
If a self selected sample was used where Ps are volunteers = BIAS
If only Ps from 1 culture was used = BIAS
If student samples are used = BIAS
What study uses a self selected sample?
Hall and Player - 70 volunteers from Met Police fingerprint Bureau who responded to a request to take part in their study
Why is using a self selected sample open to sampling bias?
The sample may not accurately represent every expert in the field because those who were confident in their ability to analyse fingerprints and make identification or those who are more thrill seeking will be sampled: not representative of experts who don’t possess these characteristics
Counter argument to self selected sample: sampling bias
Self selected samples are the most ethical because they gained informed consent if the participants volunteered. This improves the reputation of psychology making it easier to gain participants in the future
What study only uses participants from 1 culture?
Hall and Player, all participants were experts from the London Metropolitan fingerprint Bureau
Why is using Ps from 1 culture = sampling bias?
Does not measure how emotional context affects identification of prints in other police departments in different cultures who may undergo different training, so the results that the experts identification wasnt affected by emotion isnt generalisable
Counter argument to 1 culture used: sampling bias
If what is researched is species specific then using participants from 1 culture doesn’t matter. For example, the biases in fingerprint analysis use cognitive processes of selective attention to confirm expectation is a species specific aspect of cognition.
What study uses a student sample only?
Dror’s research into 27 students
Why is using a student sample sampling bias?
Because the target population of those most relevant to the research (fingerprint experts) aren’t included in the study so the results cant be generalised to how fingerprint experts may be biased by emotion irl
Counterargument to using student samples only
Student samples are most convenient for collecting a large sample, and can therefore prove a consistent effect (increasing reliability of research)
First suggestion to reduce bias in collection of forensic evidence
Use filler samples of forensic evidence alongside the evidence sample
Implementation of filler samples
Present expert with suspect’s evidence sample and 5 filler samples, ensuring the expert doesnt know which sample is which
So during analysis, expert will be more objective and less likely to claim to find an identification
Rationale behind filler samples
Miller suggested this ‘6 pack idea’ which should reduce need determination bias where expert will be more likely to find an identification because they feel its necessary. So, if they know some samples are filler, theyll be more likely to be careful in analysis
Second suggestion to reduce bias in forensic evidence analysis
Limit all information about the crime including its context thats available to them
Implementation of limiting information
Experts should only be given the suspect print and evidence print and all information on nature of crime, victim information and suspect information should be kept confidential by the rest of the police force
Rationale behind limiting information in forensic evidence
Dror’s research with students found participants who knew there was a high emotional context behind crime were more likely to find an identification of fingerprint matches, as well as when subliminal messages were given (66% matches made in this case). Due to participants being aware of high emotional context through images
Final suggestion to reduce bias in analysis of forensic evidence
Educate themselves and be consciously aware of cognitive biases within top down analysis of fingerprints
Implementation of educating themself on cognitive biases
Understand the following biases by completing a course in psychology during training: conformity effect where they are more likely to agree with identification by peers, observer bias where their prior expectation for identification influences their analysis. Train themselves to use more bottom up approaches of data driven analysis of ridges etc
Rationale behind educating themselves on cognitive biases
Hopefully, participants awareness of these biases will ensure they make a conscious effort to not repeat them, as Dror claimed these biases in top down processes heavily bias identification of forensic evidence, since the analysis is based on expert’s prior experience and expectation than objective fact
Findings of Hall and Player
57/70 participants actually read the crime scene report and knew the emotional context
13/70 didnt
6% of participants in LE context thought they were biased vs 52% of Ps in HE context thought they were biased
