Ongoing and Potential Lines of Research in Steve Lindsay’s Lab at UVic
Liz Brimacombe and I have an ongoing series of studies exploring how those in the role of police investigators weigh various kinds of forensic evidence, with particular attention to eyewitness identification evidence. In the studies conducted to date, subjects have been undergraduates. Subjects are tested in pairs, with one role playing being a police officer and the other a witness. The subject/witness watches a video of a crime, then is interviewed about that crime by the investigator interview a witness, who then searches a data base of potential perpetrators (which includes info about the appearance, prior arrest record, alibi, etc., of each of a number of individuals), chooses a suspect from that data base, make various ratings about their suspect (e.g., the probability that their suspect is the perpetrator), administer a lineup ID test to the witness, and once again making various ratings of their suspect and of the witness. Our central interest is in how the witness’s ID judgment affects the investigator’s ratings. Our results indicate that undergraduate subject/investigators are greatly influenced by subject/witnesses, regardless of the latter’s accuracy. A few years ago we developed and test-drove a new set of materials for this project and tested a small number of police officers as subject/investigators. I would like to get back to that project. This is a big project, not for the faint-hearted, and it has been on hold for a fair while. One advantage is that it is quite well defined. The main challenge is recruiting police participants.
The Effect (?) of an Advanced Warning on Eyewitness Identification Line-up Test Performance
There is a large literature on eyewitnesses’ ability (or lack thereof) to identify a culprit from a lineup (usually using a photospread lineup). Some studies have used richly realistic staged crimes, although most instead involve having witnesses watch a videotape of a simulated crime. Either way, researchers usually don’t warn subjects before they watch the video that they are going to see a crime and be asked to do an ID. Instead, subjects are given some cover story about why they are watching the video, so that the crime is a surprise and so is the lineup task. The rationale is that usually people don’t expect a crime or anticipate a lineup. BUT ethically it would be desirable to tell people in advance so that they can give truly informed consent. Also, there are real-world situations in which the chances of witnessing a crime are high and in which witnesses can reasonably be expected to know that they will have to take a lineup test (e.g., people who work the night shift at a convenience store in a tough neighbourhood). As far as I know, there is no published comparison of lineup performance of warned vs. nonwarned people. I did hear some conference presentations years ago on attempts to train store clerks to get better at attending to faces, and those attempts failed. So the idea is simply to compare the performance of subject/witnesses who are versus are not warned in advance. One down side of this project is that my bet is that it would yield a null effect, but provided the data provide strong support for the null that would be interesting.
Relative versus Absolute Judgments in Eyewitness Identification.
The Effect (?) of Feedback on Target-absent Practice Lineups on Responses to a Final “Real” Lineup
Former graduate student Mario Baldassar and I conducted a bunch of studies on this. Subjects witnessed a staged crime, then watched 5 videos and took a 5 Target-absent (TA) lineup tests (one for each crime video), with half of the subjects getting feedback as they go along, then take a TP or a TA followed by a TP lineup for the staged crime. The main question was whether getting feedback on the practice tests lowers false alarm rates on the final test. The overwhelming message was No.
The Materials-Based Bias Effect in Recognition Memory
Justin Kantner and I stumbled into the finding that when people take a recognition memory test (i.e., having been exposed to a bunch of study items, they are then given a test in which studied items are mixed with new items and asked to say which ones they previously studied), they tend to be much more conservative (i.e., less inclined to say “Yes, I studied that one”) if the materials are high-resolution scans of masterwork paintings than if they are humdrum words. This is a very robust effect, and graduate student Kaitlyn Fallow and I have ongoing series of studies aimed at understanding what gives rise to the effect.
Led by former graduate student Tanjeem Azad, my lab has conducted a number of studies in which pairs of subjects sit together and watch a video on a special backprojected screen that uses polarizing filters so that one subject sees one version of the video and the other sees a slightly different version. We then have the two witnesses talk, such that they are likely to contaminate one another with their different views. Later, we test them individually, looking for evidence of such contamination. A very fun line with lots of potential.
Again with Tanjeem Azad, in this research we expose subjects to verbal suggestions to the effect that certain details clearly presented in a video were not shown in that video. For example, the video clearly showed that the thief was wearing a cap with an Adidas logo, but two “witnesses” report that no logo was visible in the logo. We then test for acceptance of this suggestion and try to tease apart the nature/mechanisms of such acceptance.
Effects of Photos on Judgments
With Maryanne Garry and Eryn Newman (both of New Zealand) and my graduate student Justin Kantner, I have been involved in studies of the effects of marginally relevant photos on people’s judgments. In several studies, we asked subjects to judge yes or no whether or not each of a long list of mildly famous people (e.g., Jimmy Connors) is still alive. People were more likely to say “Yes, still alive” if the name was accompanied by a photo of the person (e.g., Jimmy Connors playing tennis, which is what he was famous for some decades ago). More interestingly, other subjects were given the same materials but asked to say yes or no whether the person was dead. They were more likely to say “Yes, dead” if the name was accompanied by a photo. This is a fun puzzle and we have various follow-up studies underway.
Autobiographical Memory For Scars
I have a fair-sized scar on my left forearm. I don’t remember how or when I got it. The event must have been at least somewhat traumatic. What the heck was it? I bet lots of middle-aged people have scars whose provenance they don’t recall. It would be interesting to know if performance on a remember-your-scars test correlated with other measures (e.g., repressive coping style or dissociative experiences scale). Eryn Newman and I have data from a couple of projects along these lines that need additional exploration, thinking, and writing.
Memory for Weddings
At this stage this is just a half-baked idea of using wedding videos to test hypotheses regarding memory for important events. Gender, delay, and estimated number of times of viewing the video (and/or photos of the wedding) come to mind as easy variables of potential interest. This would probably take too long to be a feasible honours project.