Effects of response disparity on stimulus and reinforcer control in human detection tasks.
Finer, more distinct response choices improve signal detection and cut bias in humans.
01Research in Context
What this study did
Roane et al. (2001) asked adults to spot faint signals on a screen. The team changed two things: how different the signals looked and how different the response buttons were.
People could press two, four, or eight buttons. The signals could be easy or hard to tell apart. The researchers tracked hits, misses, and bias.
What they found
When the buttons felt very different, people made fewer mistakes. Their bias also dropped.
The same happened when the signals looked more distinct. Both sharper signals and sharper actions improved detection.
How this fits with other research
Newman et al. (1991) showed pigeons need clear stimulus contrast for reinforcer ratios to work. S et al. now show humans also gain from clear response contrast.
Bowe et al. (1983) found pigeons look more when stimulus salience is high. The new study adds that response salience matters too.
McConkey et al. (1999) added extra choice keys without hurting pigeon accuracy. S et al. show that, for humans, more distinct keys actually help accuracy.
Why it matters
You can sharpen discrimination by giving clearer response options. Try spacing icons farther apart, using different textures, or adding color cues on AAC buttons. Even if the pictures stay fuzzy, distinct movements cut bias and errors.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Space your response options farther apart or give each a unique texture.
02At a glance
03Original abstract
In two detection experiments, university students reported whether the second of two sequentially presented tones was longer or shorter than the first by responding to stimuli presented on a touch screen. Stimulus disparity and response disparity were manipulated to compare their effects on measures of discrimination and response bias when the reinforcement ratio for correct responses was asymmetric. Choice stimuli consisted of squares filled with different pixel densities. Response disparity was manipulated by varying the difference in density between the two choice stimuli. In both experiments, decreasing stimulus disparity reduced discrimination but had no consistent effect on bias. Decreasing response disparity also reduced discrimination in both experiments, and often reduced estimates of bias. The effects of response disparity on bias were most clear in Experiment 2, in which a greater overall level of response disparity was arranged. The data show that, like corresponding research with pigeons, detection performance of human subjects can be conceptualized as discriminated operants.
Journal of the experimental analysis of behavior, 2001 · doi:10.1901/jeab.2001.75-183