Rick Stone and Dr. Brock Kirwan, Department of Psychology
This project has been worked on for nearly a year and a half. It began because research done by others indicated that there was greater hippocampal activity for an associative memory task than for a non-associative memory task. We believe that the greater hippocampal activity they observed may have been a result of higher interference, rather than higher associativity, so we tried to design an experiment to distinguish between the two.
In task 1 (single item), subjects were shown words with either a red or green background and asked to imagine each word in the color of the background. In task 2 (high interference, high associativity), subjects were shown words with either a red or green background and asked to imagine each word with a stop sign if shown with a red background, or a dollar bill if shown with a green background. These instructions remained the same for all designs we tried. Task 3 (low interference, high associativity) had several changes, described below.
In Spring Term 2011, subjects were shown words with one of eight different colored backgrounds. They were trained beforehand on what object each color represented, and told to imagine each word with the object represented by the color of the background. Subjects were tested after all three conditions had been studied.
In Fall Semester 2011, subjects studied conditions 1 and 2, and were tested after each condition, rather than being tested after all tasks had been completed.
In Winter Semester 2012, another version of task 3 was added. Instead of using eight new colors, subjects were only shown words with red or green backgrounds. Red would represent one of four objects, and green would represent one of four objects, divided into four parts. Before each of the four parts of task 3, subjects were told what object was represented by red and what object was represented by green; they were then instructed to imagine each word with the object represented by the color of the background. Subjects were tested after studying each condition (once after task 1, once after task 2, once after task 3).
Later in Winter Semester 2012, task 3 remained the same, but the method of testing was altered. Instead of testing subjects on all words in one test, the test was split into four parts, and subjects were told for each part of the test which words they were being tested on, and which objects red and green had represented for that set of words. This change was made in an effort to reduce interference.
In March 2012, after considering the results of all experimental designs, we determined that our new task had not been measuring what we hoped to measure. We were aiming for a condition that would have high associativity and less overlap than condition 2. However, our results showed that subjects had poorer memory for the third condition than for the second, suggesting that there was higher interference in task 3 than there was in task 2 rather than lower interference. We were unable to think of any other designs that we had not already tried which would provide a new
condition with high associativity, but lower interference. Before this project can be continued, a task will need to be designed that meets these criteria.