Human Information Processing Systems

This work is licensed under a Creative Commons License.

Human Information Processing Systems

The problem solving process can be considered a human information processing system. Haber (1969) has pointed out some basic assumptions of the information-processing approach in psychology. His assumptions are called (1) the stage assumption and (2) the limited capacity assumption. It is assumed in the stage assumption that the processing of information can be broken down into sub processes or stages. That is, the time interval between the stimulus and the response can be divided up into smaller intervals, and each of these corresponds to some subset of events that intervenes between stimulus and response. In the course of the stages the original information undergoes successive transformations. For example, a visual event is changed first to a recognized category (say, a red traffic light) and next to a condition for applying a rule (for example, I had better stop this vehicle). In essence, isolating a stage of information processing is not done arbitrarily: rather, a stage of processing generally corresponds to some representation of the stimulus information. As the information goes from one stage to another, its representation changes accordingly.

At each stage of processing in the limited capacity assumption, we can identify limits on the human capacity to process information. For example, if we are driving and see a red light, a traffic cop, several pretty.pedestrians, and a fire truck, then we might have too many stimuli to register in the visual system at the same time. This results in a sensory overload on capacity, and overloads can lead to several complications. First, some information may not enter the system. Second, we might recode the stimulus situation; that is we might transform it to a new stimulus. Another option might be to process the information more selectively; we might just look at the pedestrians, and ignore the light, cop, and fire truck.

Newell and Simon (1972) further define several different components that make up the human information processing system (IPS):

1. There exists a set of elements called symbols.
2. A symbol structure consists of a set of tokens (equivalently, instances or occurrences) of symbols connected by a set of relations.
3. Memory is a component of the information processing system, capable of storing and retaining symbolic structures.
4. An information process is a process that has symbol structures for some of its inputs and outputs.
5. A processor is a component of an IPS consisting of:

a. a fixed set of elementary information processes (eip’s)
b. a short-term memory (STM) that holds the input and output symbol structures of the eip’s;
c. an interpreter that determines the sequence of eip’s to be executed by the IPS as a function of the symbol structure in STM.

6. A symbol structure designates an object if there exist information processes that recognize the symbol structure as input and either:

a. affect the object directly; or
b. produce as output, symbol structures that depend on the object.

7. A symbol structure is a program if:

a. the object it designates is an information process and
b. the interpreter, if given the program, can execute the designated process.

8. A symbol is primitive if its designation is fixed by the elementary information processes or by the external environment of the IPS.

With the components of the IPS defined, we can now study the actual system. A commonly accepted and simplified model is presented below.

In the first stage of information processing, a certain amount of information about the stimulus is registered or entered in the system. The holding place for information is referred to as the sensory register because information enters the system by one or more of the five senses and is held briefly in sensory form. The information can stay in a register for a brief time, but the longer it stays there the weaker it gets (decays).

While information is in a sensory register, the processor comes into play. One of the characteristics of the processor is pattern recognition, a complex process involving contact between the information in a sensory register and previously acquired knowledge. That is, a pattern is said to be recognized when the sensory aspects of the pattern are in some way equated with meaningful concepts. Thus, pattern recognition can be thought of as assigning meaning to a stimulus. It serves the function of briefly holding information about a stimulus in the system in what is called “veridical” form—that is, in much the same form as was initially presented--until it can be put into a new form and sent further into the system.

Closely related to pattern recognition is the process of attention. One definition of attention is “paying attention to” in the sense of tuning in, not tuning out. Selective attention makes it possible to focus on, or tune in, the relevant information and to filter out the rest. Thus attention ensures that the more important information is brought into the limited-capacity system.

The smallest units of information held in the memories of the information processing system are symbols. There is no evidence that human long term memory (LTM) is fillable, and, hence, an infinite vocabulary of symbols may exist.

Human memory is described as being associative. Associativity is achieved by storing information in LTM in symbol structures, each consisting of a set of symbols connected by relations. Through learning, certain stimuli or patterns of stimuli from the input channels come to be designated by particular symbols and become recognizable. These recognizable stimulus patterns are called chunks. These stored symbols then serve as the internal representation for the corresponding stimulus patterns or chunks, and the chunks, on recognition, evoke their stored designators. These chunks are not innate, but are learned.

The IPS has a short-term memory of very small capacity. It appears that the contents of STM at any given moment consist of a small set of symbols, each of which can designate an entire structure of arbitrary size and complexity in LTM. The STM seems to be immediately and completely available to the IPS processes. The STM can be defined functionally as comprising the set of symbols that are available to an IPS process at a given instant of time (Newell and Simon, 1972).

According to Newell and Simon (1972), the amount of processing that can be accomplished in the IPS per unit of time depends on three parameters: (1) the number of processes it can do simultaneously, (2) the time it takes to do each process, and (3) the amount of work done by each individual process.

The human IPS is basically a serial system, in that it can only execute one elementary process at a time. Assuming an IPS to be serial does not imply that it cannot be aware of many things at once in the environment, in the sense of detecting and recognizing when a single stimulus occurs. This would seem to imply that behavior is serial if the information produced by a first process is required as input to a second. Hence, a serial IPS is one that can execute a single elementary process at a time.

The second factor determining how much processing the IPS can do is the time it takes to execute each information process. For elementary processes, memory access time provides an ultimate lower bound on processing time, for a process either takes as its input symbols already in STM or gets the inputs from another memory (either LTM or external environment)

The final factor that determines the speed of processing is how much work each process accomplishes. This subject will be covered in greater detail in the section on “Searching the Problem Space.”

Top