formats

What are mental codes?

Consider what has to go on in your mind in order for you to carry out the instructions for a typical Choice Reaction Time task, such as: “press the left key when you see the color green and the right key when you see the color blue.”

When a stimulus appears, at least three things have to happen: 1) you have to figure out the color of the stimulus, 2) you have to decide which key to press, and 3) you have to actually press the key. These are usually considered the three most basic, broadly-defined processes involved in carrying out a task, and are usually called stimulus identification, response selection, and motor programming, respectively (although they can be broken down into more specific sub-processes, as well; see Sanders, 1980, 1990).

These processes can be described more concretely in terms of information and mental codes. Your senses give you signals that contain information about what is going on in the world around you. In order to understand and react to the world, you use that information to create a mental picture of your environment. In other words, you form a stimulus code: a mental representation of what properties are in the stimulus environment that produced the sensory signals that you received.

Stimulus identification can be thought of as the process of forming stimulus codes based on sensory information. Those stimulus codes, in turn, contain information that can be used to decide on a response. In order to act on the world, you form a response code: a mental representation of what actions you want to carry out. Response selection can be thought of as the process of forming response codes based on information from stimulus codes. Finally, motor programming is the process of using information from response codes t prepare specific muscular movements that carry out your response. This can be thought of as the formation of motor codes, which are the programs your muscles use to make a response.

The idea that thought and action in the world consists of mental codes (representations of stimulus properties and response actions) is called the information processing approach, and models of performance based on this framework are called information processing models (see Anderson 1995; Bower 1975; Miller, 1988). Performing a task requires transforming information from the world into a stimulus code, a response code, and then a motor code, through a sequence of mental processes.

These processes clearly depend on one another. In the example above, what key you press depends on what side (left or right) you decide is correct, and what side you decide is correct depends on what you think the color of the stimulus is. In the language of information processing models, the output of stimulus identification, which contains information about the stimulus code, is used as the input for response selection. Similarly, the output of response selection, which contains information about the response code, is used as input for motor execution. Information processing models use terms like “input” and “output” a lot, because they were originally motivated by the idea that mental processes are like computer programs, and mental codes are like computer data (see Newell, Rosenbleem, & Laird, 1989; Simon, 1981; Simon & Kaplan, 1989).

Psychologists want to know exactly what is going on in these processes; that is, how information is represented in these codes, and how they are actually formed. One way to approach this question is to measure people’s performance, their speed and accuracy when carrying out a task, under different kinds of task conditions. The amount of time it takes for you to make a response is related to how difficult each of these processes is: when something about the task makes your response faster or slower, it is because one (or more) of these processes has been helped or hindered. By examining how different kinds of task conditions influence performance, psychologists are able to get an idea about what is actually going on in the formation of these different mental codes. This approach is called mental chronometry (see Meyer et al., 1988; Sanders, 1993).

There are a number of specific questions one can ask about the formation of mental codes during choice reaction time tasks. Is input information compared to items in memory one by one, until a match is found? Or is the input information compared to all possible items in memory at once? Does input information for a process cause mental codes to form gradually, or do mental codes form in chunks, like “yes” and “no” decisions? Does incomplete information get used by later processes, or do they have to wait until the previous process is completed?

Even more questions can be asked about consistency effects in classification tasks. How does irrelevant information affect the formation of mental codes? Does it influence the formation of stimulus codes, response codes, or motor codes? Does irrelevant information always have the same kind of influence on mental codes, or does it depend on task conditions?

Today, most models of consistency effects share a few basic assumptions about mental codes and how they behave during classification tasks (see, e.g., Barber & O’Leary, 1997; Kornblum et al., 1990; O’Leary & Barber, 1993; Lu & Procter, 1995; Prinz, 1990; Proctor, Reeve, & van Zandt, 1992; Umilta & Nicoletti, 1990; Wallace, 1971). For example, they assume that irrelevant stimulus codes form automatically, that different stimulus features are formed by multiple parallel identification processes, that mental codes are abstract representations, and that mental codes form gradually over time.

However, they also disagree on a few very key assumptions about mental processing. For example, different models often disagree about where selective inhibition happens. They also can disagree about whether the formation of response codes from stimulus information is continuous or happens only in discrete stage-like chunks. Finally, they can disagree about whether irrelevant stimulus information influences the formation of stimulus codes, the formation of response codes, or both.

This last question is the key issue that differentiates the Dimensional Overlap Model from other models of consistency effects. Most models of consistency effects assume that irrelevant stimulus information influences the formation of response codes, whereas the Dimensional Overlap Model assumes that influence of the irrelevant stimulus depends on what the irrelevant stimulus dimension overlaps with.

 
 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn