Individual differences, aging, and IQ in two-choice tasks

Department of Psychology, The Ohio State University, Columbus, OH 43210, United States.
Cognitive Psychology (Impact Factor: 5.06). 12/2009; 60(3):127-57. DOI: 10.1016/j.cogpsych.2009.09.001
Source: PubMed


The effects of aging and IQ on performance were examined in three two-choice tasks: numerosity discrimination, recognition memory, and lexical decision. The experimental data, accuracy, correct and error response times, and response time distributions, were well explained by Ratcliff's (1978) diffusion model. The components of processing identified by the model were compared across levels of IQ (ranging from 83 to 146) and age (college students, 60-74, and 75-90 year olds). Declines in performance with age were not significantly different for low compared to high IQ subjects. IQ but not age had large effects on the quality of the evidence that was obtained from a stimulus or memory, that is, the evidence upon which decisions were based. Applying the model to individual subjects, the components of processing identified by the model for individuals correlated across tasks. In addition, the model's predictions and the data were examined for the "worst performance rule", the finding that age and IQ have larger effects on slower responses than faster responses.

Download full-text


Available from: Anjali Thapar, Feb 05, 2015
  • Source
    • "The importance of this separation is illustrated by applications of the diffusion model in aging research. Ratcliff et al. (2001, 2003), Ratcliff, Gomez et al. (2004), Ratcliff, Thapar, and McKoon (2007, 2010, 2011) have found that the usual aging effect – slower responses for older adults – often comes about not because the quality of the information they obtain from stimuli is less (i.e., not because their drift rates are lower) but instead because their nondecision component is slower and because they set more conservative boundaries, requiring more information to be accumulated before executing a response (e.g., Starns & Ratcliff, 2010). Thus, the frequently stated conclusion that older adults' cognitive processes are, overall, worse than young adults' because all cognitive processes are slowed is incorrect. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Millions of adults in the United States lack the necessary literacy skills for most living wage jobs. For students from adult learning classes, we used a lexical decision task to measure their knowledge of words and we used a decision-making model (Ratcliff's, 1978, diffusion model) to abstract the mechanisms underlying their performance from their RTs and accuracy. We also collected scores for each participant on standardized IQ tests and standardized reading tests used commonly in the education literature. We found significant correlations between the model's estimates of the strengths with which words are represented in memory and scores for some of the standardized tests but not others. The findings point to the feasibility and utility of combining a test of word knowledge, lexical decision, that is well-established in psycholinguistic research, a decision-making model that supplies information about underlying mechanisms, and standardized tests. The goal for future research is to use this combination of approaches to understand better how basic processes relate to standardized tests with the eventual aim of understanding what these tests are measuring and what the specific difficulties are for individual, low-literacy adults.
    Preview · Article · Nov 2015 · Cognition
  • Source
    • "Average data, often accounting for a non-accurate representation of the individual subjects, have been used in the RDM fitting quite successfully. Specifically, in more than a dozen large studies with 30 to 40 subjects per group, all parameter values obtained from fitting the model to data averaged over subjects were found close (within 1 or 2 SD of each other) to the parameter values obtained from averaging the parameters resulting from fits of the model to the data from individual subjects [51] [52] [53]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: We are examining whether robust behavioral laws, initially designed to describe sensorimotor control of the upper extremity, can also describe lower extremity. Herein, we present our initial results of our research on measuring ankle reaction time (RT). We show that RT measured in ankle dorsiflexion and inversion-eversion of 7 healthy young subjects followed a γ distribution, a typical finding in upper limb response modalities. We used the mean and variance of the best fit γ distribution to find subjects with similar responses. We propose that this approach to find subjects with similar responses affords a rationale for clustering subjects with similar response to form super-subjects (SS), i.e., subjects with RTs accumulated across similar subjects. We then show that the most widely used model of RT cognitive processes, the Ratcliff diffusion model, is adequate to describe ankle RT in an SS. Results indicate that RT shows a great potential to be used as an assessment tool for our adaptive assist-as-needed robotic therapy delivered to the lower limbs of children with Cerebral Palsy.
    Full-text · Conference Paper · Aug 2014
  • Source
    • "Today however, many researchers agree that elderly people slow down, at least in part, because they choose to be more cautious than young people. This means that elderly participants often choose to collect more evidence before they are willing to commit to a decision, a strategy that may result in a substantial loss of speed in return for a small gain in accuracy (Salthouse, 1979; Starns & Ratcliff, 2010; Strayer, Wickens, & Braune, 1987; Ratcliff, This research was supported by Veni and Vidi grants from the Dutch Organization for Scientific Research (NWO). We thank Roger Ratcliff for providing us with the data that are analyzed in Experiment 2. "
    [Show abstract] [Hide abstract]
    ABSTRACT: People generally slow down after they make an error, a phenomenon that is more pronounced for older individuals than it is for young individuals. Here, we examine the origin of this age-related difference in posterror slowing (PES) by applying the diffusion model to data from young and older participants performing a random dot motion task and a lexical decision task. Results show that the PES effects on response time and accuracy were qualitatively different for young and older participants. A diffusion model analysis revealed that following an error, older participants became more cautious, processed information less effectively, and spent more time on irrelevant processes. This pattern was evident in both the random dot motion task and the lexical decision task. For young participants, in contrast, the origin of the PES effect depended on the task that was performed: In the random dot motion task, the PES effect was due to time spent on irrelevant processes; in the lexical decision task, the PES effect was due to increased caution and decreased effectiveness in information processing. Overall, PES effects were much larger in the lexical decision task than in the random dot motion task. These findings indicate that PES originates from the interplay of different psychological processes whose contribution depends on both task settings and individual differences. (PsycINFO Database Record (c) 2012 APA, all rights reserved).
    Full-text · Article · Sep 2012 · Psychology and Aging
Show more