Purpose: In this project, researchers explored the underlying cognitive processes adults with low reading skills using during comprehension. They examined the component processes adult readers use as they occur in real time, comparing the patterns exhibited by struggling adult readers with non-struggling, college-aged and older readers. By determining the connection among these processes and reading comprehension, the researchers aimed to improve models of reading for adults with low literacy to lead to improved interventions and assessments.

Project Activities: The research team used methods common in cognitive psychology and psycholinguistics (such as using verification tasks and reaction time data) to investigate reading comprehension and memory of adult readers with the aim of developing reading skill profiles for participant subgroups. These profiles included factors such as reading level, vocabulary level, and IQ. Researchers conducted studies focused on reading comprehension and other studies on how memory representations for during reading comprehension. For example, the team explored the interaction between vocabulary knowledge and the ability to recognize or make links between words and subsequent memory for and reasoning about the text. The studies included adults who were struggling readers, college-age non-struggling readers, and older non-struggling readers.

Structured Abstract

THE FOLLOWING CONTENT DESCRIBES THE PROJECT AT THE TIME OF FUNDING

Setting: The research will take place  in the Columbus, Ohio area.

Sample: The sample will include adults ages 16 and older. There will be four groups of participants: students enrolled in adult basic literacy education (ABLE) classes who read at Level 3 and at Level 4 as defined by the Tests of Adult Basic Education (TABE), college students, and older adults (ages 60 to 75).

Intervention: Proficient readers are able to decode words, locate words in memory, and create stable and accurate mental representations of texts (e.g., the ability to link referents together, use prior knowledge, and make inferences). An individual reader's ability to perform each of these processes may vary depending on characteristics such as age, reading level, or memory. Each of these processes can be tested and tracked in real-time and used to create profiles of the readers' skills. These profiles can, in turn, be used to determine which interventions are most likely to improve learner outcomes and which assessments are most useful for determining what sort of intervention a learner should receive.

Research Design and Methods: The researchers will conduct a series of 13 total studies: 1 memory, 2 lexical decision, and 10 focusing on comprehension studies in order to identify where processing difficulties lie. The purpose of this work is to work out where the bottlenecks are in processing textual information, in representing the structure of the information, in generating appropriate inferences from the information, and in remembering the information. The comprehension studies include two studies that focus on sentences, two that focus on short paragraphs, one that focuses on longer paragraphs, two that study anaphor resolution, two that study the ability to link general knowledge with texts, and one that studies forward inference. For each of these studies, the researchers will collect data on a participant's speed and accuracy and will create individual profiles based on the participant's speed-accuracy trade-off preference. The studies will use counter-balanced, within-subjects designs. The researchers will also collect data on participants' vocabulary level, IQ, and word decoding skills.

Control Condition: Given the within-subject design of the exploratory studies, each participant will serve as his or her own control.

Key Measures: In each reading study, the key measures are the time it takes a participant to respond to simple test items about a text that they are reading or attempting to remember and the accuracy of the responses. These data are input to a computational model that produces measures of the degree to which a participant understands a text and remembers it. In addition, the researchers will also collect measures of each participants' vocabulary (the WAIS vocabulary test), IQ (the combination of the WAIS vocabulary and matrix reasoning tests), and word decoding skills (the TOWRE tests).

Data Analytic Strategy: For all the studies and all the measures, differences among groups of individuals defined by IQ, reading level, age, TOWRE scores, and demographic variables such as age at which English was first spoken will be compared. Researchers will use analyses of variance (ANOVA) to determine significant differences among the individuals and groups of individuals. ANOVAs will be used to test differences in response times, accuracy, and diffusion model components. Analyses will also include explorations of the correlations among the component processes.

Products and Publications

ERIC Citations: Find available citations in ERIC for this award here.

Journal articles

McKoon, G., & Ratcliff, R. (2017). Adults with poor reading skills and the inferences they make during reading. Scientific Studies of Reading, 21(4): 292–309.

McKoon, G., & Ratcliff, R. (2016). Adults with poor reading skills: How lexical knowledge interacts with scores on standardized reading comprehension tests. Cognition, 146: 453–469. Full text

Ratcliff, R., Huang-Pollock, C., & McKoon, G. (2018). Modeling individual differences in the go/no-go task with a diffusion model. Decision, 5(1), 42–62.

Ratcliff, R., Smith, P. L., & McKoon, G. (2015). Modeling regularities in response time and accuracy data with the diffusion model. Current Directions in Psychological Science, 24(6): 458–470.

Ratcliff, R., Smith, P.L., Brown, S.D., & McKoon, G. (2016). Diffusion decision model: Current issues and history. Trends in Cognitive Sciences, 20(4): 260–281.

Ratcliff, R., Thompson, C.A., & McKoon, G. (2015). Modeling individual differences in response time and accuracy in numeracy. Cognition, 137: 115–136.

Smith, P.L., Ratcliff, R., & McKoon, G. (2014). The diffusion model is not a deterministic growth model: Comment on Jones and Dzhafarov. Psychological Review, 121(4): 679–688.