UHCL Faculty Research Briefs – April 2017

April 27, 2017 | Jim Townsend

Sometimes, your GPS doesn’t know any better

Has this ever happened to you? You are driving along when the tinny voice coming out of your navigation system takes you the long way around to your destination, even though you know the shorter route. But you drive as you are told, trusting that your electronic companion knows more than you do. University of Houston-Clear Lake Assistant Professor of Psychology Steven C. Sutherland investigates this dependence with colleagues in the article “Effects of the Adviser and Environment on Requesting and Complying with Automated Advice.”

“Given the rapid technological advances in our society and the increase in artificial and automated advisers with whom we interact on a daily basis, it is becoming increasingly necessary to understand how users interact with and why they choose to request and follow advice from these types of advisers,” says Steven C. Sutherland, assistant professor of Psychology at the University of Houston-Clear Lake. “More specifically, it is necessary to understand errors in advice utilization.”

The article was published by the Association for Computing Machinery in the periodical ACM Transactions on Interactive Intelligent Systems: Special Issue on Human Interaction with Artificial Advice Givers.

Co-authors include Northeastern University’s Casper Harteveld, assistant professor of Game Design, and Michael E. Young, who heads the Department of Psychological Sciences at Kansas State University. They propose a framework for studying interactions between automated or artificial advisers and the humans who interact with them.

They considered several factors as variables, notably the environment, the cost (in money, time, or effort) of the advice and the reliability of the digital adviser.

“We found that less predictable environments, more reliable advisers and lower costs for advice led to overutilization, whereas more predictable environments and less reliable advisers led to underutilization,” Sutherland wrote. “Moreover, once advice was received, users took longer to make a final decision, suggesting less confidence and trust in the adviser when the reliability of the adviser was lower, the environment was less predictable, and the advice was not consistent with the environmental cues.

 “These results contribute to a more complete understanding of advice utilization and trust in advisers.”

To find out more about UHCL’s Department of Psychology, visit www.uhcl.edu/human-sciences-humanities/departments/psychology.

Using game theory to model power grid stakeholders’ resiliency

When weather brings disaster, one of the vulnerable victims is the power grid, says University of Houston-Clear Lake Assistant Professor of Engineering Management Xiaojun (Gene) Shan.

 Several factors exacerbate that vulnerability, he points out. The grid becomes more fragile as more homes, businesses and industries increase power demand. Moreover, without financial commitments to long-term investments, power failures can multiply.

“It is imperative to improve the resiliency of the electric power grid before the next extreme weather event occurs,” Shan writes in the article, “Game-theoretic Model for Electric Distribution Resiliency/Reliability from a Multiple Stakeholder Perspective,” for the journal, IISE Transactions. Resiliency can be described as “the ability to rebound quickly to the original state given a major interruption in the electric power grid,” Shan says.

He and co-authors Frank A. Felder and David W. Coit use game theory to model the interactions between private and public commitments to investment in electricity infrastructure. Financial commitment to resiliency investment directly affects the grid’s ability to rebound after a disaster, they point out.

They modeled the interactions between independent, investor-owned facilities and public utilities to study whether investor-driven decisions can achieve social optimum, the point at which social benefits and social costs intersect – the “win-win.” They found in their modeling that without governance, independent stakeholders – driven by financial concerns – don’t achieve social optimum, which could have ramifications for customers of privately-owned central heat and power or CHP plants.

Felder and Coit are professors at Rutgers University. Felder is also director of the Center for Energy, Economic and Environmental Policy at Rutgers.

Study: Different cognitive skills at different reading levels

A new study of elementary school students suggests that certain cognitive abilities predict reading comprehension above and beyond basic reading skills. Furthermore, at different grade levels, students apply different cognitive abilities.

“Although the empirical relationship between general intelligence and academic achievement is well established, that between specific cognitive abilities and achievement is less so,” wrote University of Houston-Clear Lake’s Julia Englund Strait with co-authors Scott L. Decker, Alycia M. Roberts and Emma Kate Wright. Strait is a clinical assistant professor of School Psychology and director of UHCL’s Psychological Services Clinic.

They report their findings in “Cognitive Mediators of Reading Comprehension in Early Development” in the latest edition of Contemporary School Psychology, the journal of the California Association of School Psychologists.

The study investigated the relationship between specific cognitive abilities and reading comprehension across a sample of 835 children of different ethnic backgrounds in grades 1-5. It referenced a 2007 study that gave name to two different forms of intelligence: “fluid” intelligence, the ability to solve new problems, use logic in new situations and identify patterns; and “crystallized” intelligence, the ability to use learned knowledge and experience.

Results suggested that crystallized knowledge is a common cognitive contributor to reading comprehension across all the elementary grades. Other cognitive variables – specifically fluid reasoning and auditory processing – were significant in grades 1 and 2. Long-term retrieval emerged as significant in grades 3 and 4. These variables receded in importance by the fifth grade.

“Although phonics and phonological processing-related interventions continue to be appropriate for children struggling with comprehension in early elementary grades, even above and beyond the influence of basic reading skills, interventions across all grade levels should include activities for building [crystallized knowledge]-related vocabulary and background knowledge. These skills appear to be particularly important for developing reading comprehension, rather than just basic reading skills,” Strait and her co-authors said.

“Students in upper elementary and beyond may also benefit from adding explicit instruction in long-term storage and retrieval strategies more than from phonological interventions,” they said, and concluded, “Educators should keep in mind that fluid-reasoning abilities are particularly important for developing reading comprehension skills in the early grades.”

Three of the study authors are graduates of University of South Carolina, Columbia. Decker is director of the Social Psychology doctoral program there. Wright is a psychology trainee in group practice in Charlotte, N.C.  Roberts is a neuropsychology postdoctoral fellow at Rainbow Babies and Children’s Hospital, University Hospitals, Cleveland.