So, I am currently conducting my honours project for my final year of University. I am looking at the effect of the 'appearance-change instruction' (an instruction given to eyewitnesses before a lineup to make them aware that the suspect may have changed appearance since the crime, and so to not rule out those who do not immediately resemble the mental image the eyewitness has). I was looking to use signal detection theory to measure whether the instruction lowers the eyewitnesses' decision criteria. I am using a yes/no task to assess hit and false alarm rates. Is this all I need to calculate z-scores, plot a Reciever Operating Characteristic curve, calculate measures of sensitivity, and compare decision criteria between two conditions?
Yes, I guess you can fill a report with that.
The answer heavily depends on what you want to write in your report.
You should really clarify your question and provide more detail if you are looking for help. There is much expertise gathered on Researchgate and a lot of people are eager to answer your questions if they comprehend it and understand what you are aiming at.
Best
Hi Sven,
I am currently undertaking my final year honours project, for which I'm exploring the impact of the Appearance-Change instruction on eyewitness identification rates. I was hoping to use a yes/no lineup task with target presrnt/absent conditions and then compare, seeing whether the instruction creates a decision criteria shift.
However, I am not very sure the steps I would actually take post-data collection, such as how exactly to calculate the respective decision criteria.
Hi Ata,
So my participants will all watch a 'mock-crime' before being assigned to the Appearance-Change instruction present trial, or the absent trial. Then all participants will try and identify the suspects from 2 target-present and 2 target-absent lineups and will then give ratings of confidence after each trial
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Writing