Chad Dodson - US grants

This award is funded under the American Recovery and Reinvestment Act of 2009 (ARRA) (Public Law 111-5).



Confident but mistaken eyewitness testimony is one of the leading causes of wrongful convictions. Recent published research by the investigator suggests that older adults are particularly prone to make high confidence memory errors in a variety of tasks, including tests of eyewitness memory. This project investigates this age-related increase in high-confidence errors with two series of studies. The first series focuses on suggestibility errors, that is, instances in which persons assert that they have actually encountered an event that has only been suggested to them. These studies examine: 1) the generality of the age-related increase in suggestibility errors in a more ecologically-valid paradigm; 2) whether older adults can resist making high confidence suggestibility errors when the source of the suggested information has been discredited; and 3) whether distractibility at encoding is a mechanism that contributes to making errors with high confidence. The second series of studies focuses on memory for faces, as assessed with a lineup identification test. This series examines: 1) the hypothesis that older adults are prone to make high confidence false lineup identifications; and 2) whether the vulnerability to making high confidence errors persists regardless of the strength of older adults' memory for the target face (i.e., it is not a consequence of older adults' poor memory). The need for this research is particularly important given that (1) older adults will comprise an increasing proportion of the US population: from 12% of the population in 2000 to an estimated 20% in 2030; and (2) despite the aging of society and the large amount of research on eyewitness memory, relatively little research has investigated the eyewitness memory of older adults.

Everyday situations require individuals to understand another person's level of confidence in his/her memory. For example, when a spouse says, "I'm pretty sure I locked the door" do you interpret that statement as indicating that you should double check the lock or not. Interpretation of such expressions of confidence present important cognitive challenges and can be influenced by a number of different factors. The cognitive biases that can influence these interpretations of confidence and the approaches to expressing confidence that might limit the impact of these biases are important areas of research with wide-ranging applications. The current project examines these factors in the context of eyewitness identification. 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 it was intended? While a large literature exists on eyewitness confidence and (a) its relationship with identification accuracy, (b) its influence on jurors, and (c) its vulnerability to influence from post-identification feedback, very little is known about how other people understand the verbal expressions of eyewitness confidence. The outcomes of this project in terms of a more thorough understanding of the cognitive factors that influence interpretations of confidence are anticipated to have broad implications for settings, such as judicial proceedings, that involve the interpretation of another person's confidence.

This project investigates how people understand an eyewitness's verbal expression of confidence in the accuracy of his/her identification from a lineup of faces. There are three goals: first, the project examines how the particular content of an eyewitness's justification for his/her level of confidence (e.g., "I'm pretty certain it's him because I remember his chin") influences how people understand the eyewitness's confidence-statement. Second, there are a variety of cognitive biases that cause individuals to misinterpret the verbal expressions of certainty. For example, mounting research on the 'outcome-severity bias' shows that the identical verbal probability phrase (e.g., "it is somewhat likely") is interpreted as denoting a higher numerical likelihood when that phrase refers to more severe events rather than less severe events. The project considers whether eyewitness expressions of confidence are vulnerable to misunderstandings that are caused by the same kinds of cognitive biases. Third, the proposed experiments investigate whether the harmful influence of these cognitive biases can be mitigated by expressing confidence numerically instead of verbally. Overall, the project will fill a profound lack of knowledge about how people understand eyewitness confidence in a lineup identification. Moreover, the knowledge that verbal expressions of certainty are vulnerable to misinterpretations may: (a) guide police to obtain a numerical estimate of certainty from eyewitnesses; and (b) alert police, lawyers, judges and jurors to potential pitfalls about interpreting verbal expressions of certainty.

? DESCRIPTION (provided by applicant): This is an application to continue research originally started in 2001, and expanded into a longitudinal study, known as the Virginia Cognitive Aging Project (VCAP), in 2005. Over 2,300 adults 18 - 95 years old have now completed at least two longitudinal occasions, with an average of 2.7 occasions and an average time in study of 5.1 years. The research proposed in the next funding period will extend the investigation of short-term longitudinal change in a broad variety of cognitive measures, with particular emphasis on adults under the age of 80. Although previous studies have found little or no cognitive change in longitudinal comparisons involving young and middle-aged adults, this research employs three methodological innovations, variable retest intervals, measurement bursts at each occasion, and continuous recruitment of new participants, that help distinguish age effects from experience (retest) effects, and that increase sensitivity to detect change by taking into account normal short-term variability in performance. Among the primary questions to be investigated are when does normal age-related cognitive change begin, the degree to which changes in different cognitive variables are independent of one another at different periods in adulthood, the role of prior test experience on the direction and magnitude of cognitive change at different ages, the degree to which factors such as one's cognitive or physical lifestyle moderate the amount of age-related change in different cognitive abilities at various periods in adulthood, and how early can normal and pathological trajectories of cognitive aging be distinguished. Specific aims during the next grant period are to: (1) Expand the characterization of normal cognitive aging across the range from about 18 to 80 years old; (2) Extend the investigation of the role of experience effects on cognitive change; (3) Investigate the structure and nature of cognitive change across different levels of analysis and across a wide range of ages; and (4) Increase sensitivity of VCAP tests to detect early stages of cognitive pathology among VCAP participants.