Impact of computerized cognitive training on working memory, fluid intelligence, and science achievement

Research institution: Stanford University

Research title: Impact of computerized cognitive training on working memory, fluid intelligence, and science achievement

Researchers: Richard J. Shavelson, Kun Yuan, and Alicia C. Alonzo

Training program used in research: Cogmed RM

Published: Dissertation, Stanford, August 2008

Funding: This research was supported by a grant from the Wallenberg Global Network. Training program, investigator training, and technical support provided by Cogmed. No funding provided by Cogmed.

Abstract

This is abstract is from Kun Yuan’s dissertation

Working memory is responsible for temporarily maintaining and manipulating information during cognitive activity (Baddeley, 2002). It includes short-term memory and cognitive control (Engle, Tuholski, Laughlin, & Conway, 1999). Working memory has been found to be closely related to a wide range of high-level cognitive abilities and academic achievement (Gathercole, Pickering, Knight, & Stegmann, 2004; Kyllonen, 2002). Prior studies showed children with attention-deficit/hyperactivity disorder improved their working memory and fluid intelligence (e.g., the ability to understand complex relationships and solve novel problems) through computerized cognitive training in a clinical setting (Kingberg, Firssberg, & Westerberg, 2002; Klingberg et al., 2005). This research examined whether regular middle-school students would increase their working memory, fluid intelligence, and science achievement through computerized cognitive training in a public school setting.

Two randomized pretest-posttest control-group experimental studies were conducted to test two research hypotheses: 1. Computerized cognitive training would significantly improve working memory of regular middle-school students in a school setting; and 2. increase in working memory would lead to improved fluid intelligence and science achievement. In both the pilot and main studies, students were randomly assigned into the experimental or control group after being tested on working memory, fluid intelligence, and science achievement. Experimental group students were trained by RoboMemo (2005, Cogmed Cognitive Medical Systems AB, Stockholm, Sweden), a multimedia program for Windows designed to improve working memory; control group students worked on computerized science lessons (unrelated to content in science achievement test) or MegaMemo (2005, Cogmed Cognitive Medical Systems AB, Stockholm, Sweden), a placebo version of RobeMemo providing low level working memory training. Daily training time was 30-40 minutes. Trained coaches supervised the training and gave feedback regularly throughout the training. Students took the same tests at the end of training as at the beginning to examine whether experimental group students improved significantly more than did control group students on measures of working memory, fluid intelligence, and science achievement.

Auditory Number Span Test (Ekstrom, French, & Harman, 1976), Span-board (Klingberg et al., 2005), Automated Reading Span (Conway et al., 2005), and Automated Operation Span (Unsworth, Heitz, Schrock, & Engle, 2005) were used to measure working memory capacity. Raven’s Standard Progressive Matrices Plus (Raven, 1998) was administered to measure fluid intelligence. A Science Achievement Test developed by the Stanford Educational Assessment Laboratory was used to assess students’ science achievement. Their reliability coefficients ranged from 0.76 to 0.89. Fifty-two seventh- and eighth-grader in a summer school in Northern California participated in the pilot study between June and July 2005. They were trained for 13.55 days on average.

Thirty-seven seventh and eighth grade students in an intermediate middle-school in Northern California participated in the main study between January and March 2006. They were trained for 24.65 days on average. Results showed computerized cognitive training effectively improved regular students’ working memory in a school setting, with more increase in short-term memory than in cognitive control. No significant difference was observed right after the training between experimental and control group students in changes of fluid intelligence and science achievement. With more participants, enough time to transfer changes in working memory to fluid intelligence and science achievement, and more training on cognitive control, students might significantly improve their fluid intelligence and science achievement through computerized cognitive training in future studies.

This research contributed to the understanding of human abilities’ plasticity. It also provided a useful way to improve students’ cognitive capacity and possibly academic achievement, and potentially help schools meet national performance standards.