Hi,
I am doing the project Testing for Bias in a Photo Lineup, and I had three different test groups. Some groups had higher critical ratios than others, and I wanted to know if that meant that groups with higher critical ratios had more bias.
Thanks in advance!
Does it matter?
Moderators: AmyCowen, kgudger, bfinio, MadelineB, Moderators
-
bookworm0815
- Posts: 2
- Joined: Wed Oct 23, 2013 3:38 pm
- Occupation: Student:8th Grade
- Project Question: Hi- I am thinking of doing the project called Testing for Bias in a Photo Lineup. It sounds like a great idea, because it combines two subjects I am interested in (forensic sciences and psychology). However, I wanted to know what, exactly, is the main question that I would want to answer through this process? It sounds like a dumb question, but I figured I'd ask in case anyone else is wondering too. Thanks in advance:)
- Project Due Date: n/a
- Project Status: Not applicable
Re: Does it matter?
Hello. A very interesting experiment. I have been part of juries where photo lineups were a critical part of the case, and bias was definitely discussed.
A group with a higher critical ratio doesn't guarantee a bias. It only means higher confidence that the lineup is biased.
Conducting this type of experiment doesn't provide answers that are 100% guaranteed, because we are only testing a small sample (a few witnesses) to try and make an educated guess of the outcome if every single person on the planet acted as a witness in this situation, which of course is impossible. When conducting a test with only a sample, you attach a percent confidence that your test results match what would happen if every person on the planet acted as a witness. Obviously, the higher the confidence the better.
Look at the numbers presented in steps 8i and 8j of the procedures. if your critical ratio is greater than 1.96, then you are 95% confident that the witness was not chosen just due to chance (ie, 95% confident the lineup is biased). If your critical ratio is greater than 2.58, then you are 99% confident that the witness was not chosen just due to chance (ie, 99% confident the lineup is biased).
So, when discussing the differences between your groups, discuss in terms of confidence in biased / not biased v. stating conclusively that one group is more biased over another. If critical ratios for your groups are 3, 3.6, and 4, since you are 99% confident that all three groups are biased, does it matter which group is "more biased?" Probably not.
I found a thread from a few years ago that also discusses this experiment. Check it out and see if it clarifies further. The link Rajeev provided is a good explanation of how to calculate and interpret the critical ratio.
https://www.sciencebuddies.org/science- ... =27&t=5529
I hope this helps. Be sure to write back if you have further questions. Good luck!
A group with a higher critical ratio doesn't guarantee a bias. It only means higher confidence that the lineup is biased.
Conducting this type of experiment doesn't provide answers that are 100% guaranteed, because we are only testing a small sample (a few witnesses) to try and make an educated guess of the outcome if every single person on the planet acted as a witness in this situation, which of course is impossible. When conducting a test with only a sample, you attach a percent confidence that your test results match what would happen if every person on the planet acted as a witness. Obviously, the higher the confidence the better.
Look at the numbers presented in steps 8i and 8j of the procedures. if your critical ratio is greater than 1.96, then you are 95% confident that the witness was not chosen just due to chance (ie, 95% confident the lineup is biased). If your critical ratio is greater than 2.58, then you are 99% confident that the witness was not chosen just due to chance (ie, 99% confident the lineup is biased).
So, when discussing the differences between your groups, discuss in terms of confidence in biased / not biased v. stating conclusively that one group is more biased over another. If critical ratios for your groups are 3, 3.6, and 4, since you are 99% confident that all three groups are biased, does it matter which group is "more biased?" Probably not.
I found a thread from a few years ago that also discusses this experiment. Check it out and see if it clarifies further. The link Rajeev provided is a good explanation of how to calculate and interpret the critical ratio.
https://www.sciencebuddies.org/science- ... =27&t=5529
I hope this helps. Be sure to write back if you have further questions. Good luck!
Deana