The “same or different” concept is something we all learned as a child. We were shown two images and asked if they were the same or different. This activity teaches us to compare objects in the world; it introduces critical thinking and has applications in mathematics.

The same/different task has also been used in visual recognition experiments for decades and decades. A common example of this task is to show a string of four consonants. Then, a new string of consonants is displayed, and the participant must determine if those strings are the same or different.

Study string: TLFQ

Test string: PLFQ

Response: Same or Different?

Previous studies have shown that the response time (RT) for different strings depends on where the difference is in the string (from left to right). If the difference is early (leftmost) in the string, the RT will be faster than a later (rightmost) difference. RT is a function of where the difference is in the string.

However, when the two strings are the same, RT is usually faster than when there’s a difference. People are quicker to perceive that two strings are the same than perceiving a difference—even when the difference is in the leftmost position. Researchers suggest that this distinction in RT indicates that same and different responses are controlled through distinct computational processes in perception.

In 2017, Sinha and Glass showed that different responses were associated with caudate activation, and same responses were associated with hippocampus activation. Fast forward to 2023, and Arnold Glass’s team (see below) returned to further investigate response activation in the brain. They describe a series of six experiments in their Attention, Perception, & Psychophysics paper: “Response assignment influences visual recognition.”

The experiments in their paper used similar methodology. Participants saw a string of four consonants on a screen (the study string), and then they saw a single consonant (the test consonant). The task was to determine if the test consonant appeared in the study string. In these experiments, the authors varied the location of response keys—keys could be assigned with same on the left and different on the right, or vice versa.

In one fMRI experiment, the researchers recruited 22 right-handed individuals to complete the task over two sessions. In one session, the different key was on the right, and in the other session (a week apart), the same key was on the right. Based on their previous work, the primary regions of interest were the caudate and hippocampus. They found that the left hippocampus and right caudate were more active when the different responses were on the right. The right hippocampus was more active when the same responses were on the right. The left caudate activity did not differ based on response assignment.

This is an impressive new finding! It shows that the same-different response assignment to left or right keys will determine which neural systems perform the visual recognition task. The right-sided key assignment is important because it is presumed to have a stronger connection to the left hemisphere, which is dominant for language. Therefore, the verbal label on the right may dictate which approach the perceptual system follows. When the different key was on the right, the caudate and hippocampus were actively comparing the test consonant to the study string—to determine if the consonant appeared in the string. When the same key was on the right, the hippocampus was making a decision about recency/novelty—that is, to recall if the test consonant was seen recently (without directly comparing it to the study string).

The authors commented: “This finding has profound implications for how the cognitive systems in the brain are organized and provides a new methodology for isolating their specific effects.”

These implications should guide future research using same-different tasks in which RT is an important dependent variable. Not only should key assignments be counterbalanced, but responses should be considered in light of these recent findings. Furthermore, the results from the past 50+ years of same-different experiments may need to be re-evaluated. It will be interesting to learn if those previous findings are the same or different after considering this new work.

Featured Psychonomic Society article:

Kang, M., Norman, M., Becker, A., Zhou, W., Wang, T., Xuan, S., & Glass, A. (2023). Response assignment influences visual recognition. Attention, Perception, & Psychophysics, 85:1179–1198. https://doi.org/10.3758/s13414-023-02702-8