Search Here

How can a decrease in memory accuracy be minimized when people are trying to recall information or answer a series of questions over a relatively long period of time?

Simply changing the subject matter of the questions increases accuracy on longer tests, found researchers from Syracuse University, the University of South Florida, and Indiana University. 

"The simple act of testing harms memory," said Amy Criss, assistant professor of psychology in Syracuse University's College of Arts and Sciences. "Previous studies have shown that people are more accurate in their responses to questions at the beginning of a test than they are at the end of a test. This is called output interference. Our study demonstrates how to minimize the effects of output interference."   

In the study, the researchers had participants memorize word sets from different categories, such as animal and geographic terms, or countries and professions. The testers were then split into three groups, each of which responded to a series of 150 questions. The tests included 75 terms from each word set. 

The first group of testers responded to questions in which the terms were randomly intermixed. A second group responded to 75 questions about one category followed by 75 questions from the second category. The third group responded to alternating blocks of five questions about each category. 

The second group outperformed its counterparts on the test. "While accuracy fell off as the test subjects neared the end of the first category of terms, the accuracy rebounded when the questions switched to the second category of terms," the lead author says. "The study demonstrates that memory improves when categories of information that people are asked to remember change." 

The results have implications for the way in which standardized and comprehensive tests are created, Criss says. "You don't want to place a lot of the same information into one section of the test. Accuracy will increase by changing the subject matter of the questions." 

The findings of the study appear in Psychological Science, a journal of the Association for Psychological Science.