User:Maarao/sandbox

SCHOOL ASSIGNMENT: WILL TAKE DOWN SOON

Semantic priming refers to the improvement in speed or accuracy in responding to a stimulus such as a word or a picture, when it is preceded by a semantically related stimulus (e.g. Doctor-nurse) relative to when it is preceded by a semantically unrelated stimulus (e.g. chair-doctor).

History
The term was coined in 1971 during an experiment conducted by David E. Meyer and Roger W. Schvaneveldt on high school students. In their experiment, the students were asked to decide whether two strings of letters presented at the same time were words or non words. It was discovered during this experiment that the participants were faster and more accurate in responding to set of words that are related semantically (related in meaning). From this experiment, semantic priming became a type of priming and has become very influential in the field of cognitive science. The preceding stimulus is referred to as the Prime while the stimulus to which responses are made is referred to as the target. Semantic priming caught the attention of psychologists especially in the field of cognitive psychology and cognitive science.One of major importance of semantic priming is its role in many cognitive tasks like semantic categorization and word recognition. Semantic priming also serves a role in memory, especially in retrieval of items from memory and is usually used to investigate other aspects of perception and cognition such as knowledge representation and reasoning. Semantic priming is also of great importance in studying consciousness because the priming effect is usually unconscious and automatic. A lot of research has gone into understanding this phenomenon over the years. Some researchers have focused on investigating the processes underlying semantic priming while some are investigation factors that influence this phenomenon. Some findings support that semantic priming involves a fast onset without awareness, this is known as automatic processing which supports automaticity. Other findings have however shown that semantic priming might involve strategic processing which usually has a slower onset, requires awareness and has costs and benefits. There is also some research looking at the relationship between the prime and target and some evidence suggests that the priming effect is not usually solely due to semantic relation but also depends on association. An example of this is that there would be a priming effect in a pair of words like bus and passenger. Although these two words are not semantically related, we associate them together. These distinctions have helped researchers in investigating semantic priming and the processes involved and in developing the models involved.

Research Paradigm
The most frequently used task in studying semantic priming is the lexical decision task. This task was created by David E. Meyer and Roger W. Schvaneveldt for their studies on semantic memory. The lexical decision task requires participants to respond to a string of letters (the prime and target) and determine if it is a word or not (e.g. bed vs. gret). With this task, you can measure the reaction time of each participant that is how long it takes to respond to a word versus non word and if the prime has any effect in the reaction time. This task is also of great important in investigating other forms of priming in cognitive Science. The single word paradigm is also used in studying semantic priming. It is similar to the lexical decision task but differs in the terms that participants are only required to respond to a target and ignore the prime. With the single word paradigm, the reaction time would mostly be lower if the prime and target are semantically related than if they are not. Naming or pronunciation task is also frequently used in investigating semantic priming. In this task, participants are required to read the target word aloud as quickly as possible. Just as is the case in the lexical decision task and the single word paradigm, participants usually name the target word faster when it is preceded by a semantically related word (prime) than when it is preceded by a semantically unrelated word.

Spread Activation Model
The Spread Activation Model is usually referred to as the leading model of semantic priming. It is based on the concept of spreading activation and was first incorporated into the model in 1967 by Quillian and was expanded by Allan M. Collins and Elizabeth Loftus in 1975. This model is based on the assumption that recalling an item from memory requires activation of its internal representation and that this activation spreads from one concept to another. In this model, the more related concepts are activated in memory, the easier it would be to recall the target item. The spread activation model has been used for constructing network models of memory in cognitive science. In these networks, Nodes represent concepts (e.g bus) and the lines connecting the nodes are the links which represents the relations between the concepts(e.g. a link between bus and passengers). Collins and Loftus model is a subtype of the spread activation model that differentiates knowledge of the meaning of concepts from knowledge of their names. Under the Collins and Loftus’s Model the following assumptions were made about the spreading activation process:

1) Activation begins to be released from a concept at a constant rate when it is processed and the longer the processing continues, the longer it continues to let out activation

2) Only one concept can be actively processed at any one time, but once activation is let out, its spreads in parallel in the network.

3) Activation coming to a given node from various sources summates at that node.

4) Activation takes more time to spread greater distances, with distance measured in the amount of intersected links

5) Activation decays with distance and this decay is dependent on the strength of links

6) Activation goes away/decays with time and as more mental activities get involved

According to this model, semantic priming occurs because the strength of the link between semantically related concepts are usually very strong. This means that an activation of a concept usually leads to the activations of semantically related concepts.

Becker's Verification Model
Becker’s Verification Model is based on the idea that a visually presented word is stored in out visual sensory memory and that these visually presented words have features that are analyzed by word detectors in the lexicon. For example, the letter U can be analyzed as having two lines and a curve as its features. According to this model, because many words in the lexicon will share the same features, many word detectors will be activated initially. The set of activated word detectors is referred to as a sensory set. After sampling and constructing features, a verification process is required for completing the identification of a word. The verification process samples one word from the sensory sets and builds a visual representation of the word using its initial features and related features. This visual representation is then compared to the representation of the word in the visual sensory memory. If the representations match, the stimulus is identified as the correct word usually referred to as the target. If there is no match, sampling, construction and comparison will continue until a match is found. According to this model, semantic priming occurs because the feature analysis processes are skipped. The feature analysis process is skipped because when the prime is recognized, information about the word meaning becomes available and word detectors in lexical memory are activated according to how similar in meaning they are to the prime. Word detectors that meet the criteria for activation are then used by the verification process in the same way the sensory sets are used, If the target is related is semantically (related in meaning) related to the prime, it will be recognized as the successful candidate word. In this case, recognition time will be lower for targets semantically related to the prime than those not semantically related to the prime.

Compound-Cue Models
Compound-Cue Models of priming were proposed independently by Ratcliff and Mckoon in 1988 and by Dosher and Rosedale in 1989. This model deals with the content of retrieval cues. In regards to semantic priming, the idea is that the cue to memory contains the target item and elements of the neighboring context. The compound cue model is usually combined with a model of memory to make predictions about performance in a task. For example, in a lexical decision task, the context typically includes the prime and other words occurring before the prime. According to this model, memory retrieval is a process that unfolds over time and involves many component sub processes. An important contribution of these models is that they place many priming phenomena within the view of general models of memory.

Distribution Network Models
Distributed Network Models has been around since the 13th century but only became influential in cognitive psychology in the mid-1980s. According to the distributed networks model, a concept is a pattern of activated nodes. In this model, networks learn and degrade gracefully with damage to the network. There are different categories of these models. In one of the categories referred to as Proximity Models, semantic priming occurs because related primes and targets are closer to each other in a high dimensional semantic space than primes and targets that are unrelated. In another category referred to as Learning Models, semantic priming is referred to as learning that occurs when a word is recognized or is the object of a decision. These models have been applied in studying of human behaviours including learning of novel information, word naming and impairments in reading and in linguistics.

Other Models
Other Models important in semantic priming include the Multistage Activation Models whuch includes the Logogen modeland the Interactive-Activation Model, Foster's Models,Norris's Context-Checking Model, Neely and Keefe's Hybrid Model and ROUSE. These models have played major roles in studying semantic priming.

Hemispheric asymmetries in Semantic Priming
Most language functions are usually lateralized to the left hemisphere but this is not exactly the case in semantic priming. Studies investigating hemispheric asymmetries in semantic priming have shown that there is an indication that both hemisphere contribute equally to some kinds of priming but not all. Priming that are as a result of automatic processing seems to involve both hemispheres while forms of priming that involve strategic processing seems to be lateralized in the left hemisphere.

Event Related Potention
Studies looking at semantic priming using Event-related potential have focused on the N400 (neuroscience) component. The N400 is a negative peak in potential that occurs around 400ms after presenting a stimulus. There is evidence that N400 (neuroscience) is produced by processing that involves mixing ongoing semantic events with previous sensory set. These studies have shown that the N400 component is greatly reduced and sometimes absent when the target and the prime are semantically related. The easier it is to mix features of stimuli with the sensory set, the smaller the peak of the negative potential. This phenomenon is sometimes referred to as the N400 priming effect.

Research
Most research in the area of semantic priming has focused on its relevance in word recognition and how it is impaired in various diseases (see below). Most of these studies have also looked at the effects of semantic relation, stimulus quality and repetition, and word frequency on semantic priming. Research has shown that the degree of semantic priming in lexical decision is greater for low frequency words (e.g. xylophone, pint) than for high frequency words (e.g. chair, eat, hungry). The different models of semantic priming have given researchers an idea of the processes involved in semantic priming and why this priming effect is seen. Most emphasis has been paid in distinguishing between automatic processing and strategic processing in semantic priming. Research in this area suggests that both processes are of equal importance but play different roles depending on the task used. The lexical decision task for example has been shown to involve automatic processing rather than strategic processing while processing of non familiar words seem to involve strategic processes. Most research has also shown that the priming effect occurs without the awareness of the individual. This finding has been influential in studying unconsciousness. Research has also shown that factors such as subject’s explicit expectations and semantic relation between prime and target influence the degree of the priming effect. This is no surprise as the more semantically related the prime and target are, the easier it would be for a subject to respond to the target thus there will be a bigger priming effect. There is also some evidence that priming effect diminishes with repetition of a stimulus. Some researchers have looked at the differences in semantic priming between older adults and younger adults. Research in neuroscience has shown that semantic memorytypically degrades as we age but research involving semantic priming has provided evidence that there is greater semantic priming effect in older adults compared to younger adults. This finding indicates that age is not a condition for reduced semantic processing. There has also been a lot of research on the effects of neurodegeneration on semantic priming and processes involved.

Effects of Illness
Semantic Priming has been used a tool in studying semantic processes in individuals with various neurodegenerative diseases occuring as a result of Neurodegeneration such as Parkinson’s disease and dementia. Researches investigating semantic priming in Parkinson’s patients have shown that semantic processing might be dysfunctional in these patients and that this leads to difficulty in understanding complex sentences. There is also evidence that both automatic and strategic processes are affected in these individual. There is evidence that due to disturbances in the frontal-striatal system, more attention is paid to semantically unrelated words leading to an inhibition and failure to respond to semantically related words. Schizophrenia has also been studied using semantic priming. Research in this area has focused mostly on deficits in the sensory network models of individuals suffering from schizophrenia. Studies till date have shown that these patients show variable semantic priming effects under automatic processes but semantic priming effects seem to be impaired in strategic processes. Semantic priming effect has also been studied in patients suffering from mild cognitive impairment. The results have shown that semantic priming seems to be impaired in these individuals but that this lack of priming effect is more likely due to the degeneration of semantic concept that rather than an impairment in semantic processes. This finding has led to the understanding that awareness is not impaired in these individuals and that they have no problem accessing semantic concepts in semantic memory but rather these concepts are being degraded.

Applications
Just as discussed above in the research section, Semantic priming has revealed a lot about the processes that influence word recognition. Semantic priming has also given researchers the opportunity to study the effects of stimulus quality, stimulus repetition, word frequency and semantic relation and how they interact. Research in this area has revealed that we integrate these various factors when analyzing a stimulus and that this integration leads to faster and correct recognition. For example, reading about a bus in a city where they are no buses would be harder than reading about a bus in a city where buses run. This is because bus in this case has a semantically relevant in the context and is also going to be a high frequency word thus leading to better recognition. Semantic priming has also been used in studying the degradation of semantic memory in patients suffering from illnesses that lead to neurodegeneration.

Feature Research
With the great improvements in neuroimaging techniques over the years, neural basis of processes underlying semantic priming can be studied in greater detail. It is not precise what brain areas are involved in semantic priming yet. Early studies using positron emission tomography as an imaging technique has revealed some brain areas that may be involved in semantic priming. These areas include anterior temporal lobe, anterior medial temporal, superior temporal gyrus and anterior cingulate. It was also shown that these brain areas are mostly active when the prime and target are no semantically related. More research is required in identifying if these brain areas are involved in semantic processing or in attention. Knowing the neural basis of semantic processing will help in determining what kinds of treatments would be effective in treating individuals with various illnesses such as Parkinson's disease and schizophrenia and will further understanding of the models of semantic priming.