Research Topics
Our research focuses on memory with an emphasis on the occurrence of (a) false memories, (b) overconfidence in one’s memories and (c) changes in memory across the lifespan. For the past several years, we have focused on two related issues: (a) understanding another’s expression of confidence in a memory, and (b) understanding factors that explain why some individuals are better than others at assessing the likely accuracy of a memory. We have examined both of these issues primarily in the context of eyewitness memory.
Eyewitness Memory
We have been examining factors that contribute to eyewitness identification errors, particularly those that are made with high confidence. Although growing research suggests that high confidence eyewitness identifications are generally reliable, we have focused on three factors that are systematically related to high confidence misidentifications.
One factor involves face recognition ability. Unsurprisingly, there are substantial differences from person to person in face recognition ability. We have shown (Gettleman et al., 2021; Grabman et al., 2019, 2020) that poorer face recognizers are much more likely than stronger face recognizers to make high confidence misidentifications.
Another factor is how eyewitnesses justify their identification of a face from a lineup. We (Dobolyi & Dodson, 2018; Grabman et al., 2019) show that there is a similar relationship between confidence and accuracy for identifications that are accompanied by either observable, unobservable, or recognition justifications. By contrast, there is a weaker confidence-accuracy relationship for familiarity-based identifications. One implication of these results is that an otherwise identical high confidence identification is more likely to be wrong when it is based on familiarity than when it is based on any other type of justification. These results indicate that there is great practical value for law enforcement to record eyewitness justifications and to give extra scrutiny to familiarity-based identifications.
Finally, consistent with other research, we find that highly confident identifications are more likely to be wrong when the eyewitness has taken a longer than a shorter amount of time to make the identification (Dobolyi & Dodson, 2018; Dodson & Dobolyi, 2016; Gettleman et al., 2021; Grabman et al., 2019).
Understanding Confidence: Eyewitness Testimony as a Model Case
There has been a vast amount of research on how well an individual’s confidence in a memory predicts its accuracy. For example, much research on eyewitness confidence has examined its relationship to the accuracy of an identification of a face from a lineup. Yet, apart from my lab’s work, there has been almost no attention paid to the question of how people interpret verbal expressions of eyewitness confidence. This gap in the research literature is particularly surprising because (1) growing research shows that eyewitness confidence in an identification can be a strong predictor of accuracy; and (2) people (e.g., eyewitnesses) prefer to express confidence verbally (e.g., “I’m pretty sure it’s him”) than numerically. Thus, one potential cause of judicial errors is misunderstanding an eyewitness’s intended level of confidence.
We have investigated factors and cognitive biases that cause people to misinterpret another’s verbal expression of certainty about their memory. For example, when an eyewitness identifies someone from a lineup and states, “I’m pretty sure it’s him,” how do we know that police, jurors and others will interpret this expression of confidence in the way that was intended? Perhaps an eyewitness means he is only 50% sure but others might think he is 80% sure, or vice versa.
One bias that distorts how people interpret an eyewitness’s expression of confidence is what we call the Featural Justification effect (Dodson & Dobolyi, 2015). Consider an eyewitness’s confidence statement in a lineup identification, such as “I am positive it’s him.” We have found that this identical verbal expression of confidence will be interpreted by others as indicating a lower numeric value when this expression is accompanied by a reference to a visible feature on the suspect (e.g., “I am positive it’s him. I remember his eyes.”) than when it is accompanied by a statement of recognition (e.g., “I am positive it’s him. I recognize him.”) or by a reference to an unobservable detail (e.g., “I am positive. He looks like my cousin.”; Dobolyi & Dodson, 2018; Dodson & Dobolyi, 2015; Grabman & Dodson, 2019). Moreover, eyewitnesses are perceived as less accurate and suspects are judged as less likely to be guilty when an eyewitness refers to an observable feature than to one of these other details (e.g., Dodson & Dobolyi, 2017; Slane & Dodson, 2019).
Aging & Memory
Older adults exhibit a variety of memory impairments. Although there are several different explanations of this finding, many of them share the view that older adults remember fewer specific details about past events than do younger adults. By contrast, we have proposed that older adults are prone to experience convincing illusory recollections (Dodson, Bawa & Krueger, 2007; Dodson, Bawa & Slotnick, 2007; Dodson & Krueger, 2006). But, there are constraints on this age-related increase in high-confidence memory errors: they occur on memory tests that require specific details about recently-learned events, such as remembering who-said-what earlier; they do not occur on tests of semantic memory, such as remembering who wrote the book, War and Peace. Importantly, this age-related propensity to make high confidence errors persists after ruling out alternative explanations, such as age-related differences in either overall memory accuracy or the tendency to use the high end of the confidence rating scale.