LITERATURE REVIEW

            Normally, with the increase in age, there are many changes which tend to appear to an individual. Such changes include but not limited to loss of memory and reduced processing speed of the brain. These changes have been addressed in different perspectives by different authors. With some of the research studies converging, others have indicated huge differences. Each of these studies presents their unique idea or opinions on the matter and gives supportive information for that matter.  According to Ratcliff and McKoon who investigated the relationship with aging and individual response time, they concluded that the older people were slower in response compared to the young subjects, but more accurate. On the other hand, Masson’s research study indicated that Co-occurrence can be used as a measure of ease and semantic relatedness

The primary objective that the article seeks to attain is to evaluate the compound cue Model of Ratcliff and McKoon as well as the distributed memory model of Masson. The two works by Ratcliff&Mckoon and Masson primarily dwells on the concept of automatic priming and as a result. When evaluated comprehensively, the models will be attained by carrying out a test for mediated priming, which mainly refers to the priming of the target words which are perceived to be indirectly related to the semantic memory (Chwilla & Kolk, 2002).

Every research work conducted seeks to fill a niche that was left by previous research studies. This study was motivated by research conducted by McNamara and Altarriba in 1988 and Chwilla, Kolk, and Mulder (2000) all of whom researched the same topic. The study conducted by McNamara and Altarriba (1988) noted that the presence of direct associates was one of the main reasons as to why mediated priming was inherently absent. In addition to that, they were able to demonstrate two-step priming using concrete and substantial evidence. On the other hand, the research by Chwilla, Kolk, and Mulder in 2000 questioned the hypothesis, which indicated that list composition was a crucial component in attaining two-step priming (Chwilla & Kolk, 2002). The findings in both of these studies left some questions on the topic of three-step priming unaddressed which eventually motivated this study.

As noted in the first paragraph, an evaluation is to be conducted between the two models. A researcher has to formulate a research question that seeks to be answered by the findings of the study or else develop a set of hypothesis that they will test using the results of the study to attain reliable results. For this article, the researcher stated a hypothesis which argued that if the mediated pairs were to be direct, despite the fact that they are weak and are related semantically, then this would automatically imply that the compound cue model was responsible for such effects.

To test the above hypothesis, the researcher conducted two experiments. In the experiments, 30 participants were chosen, but two were not included in the analyses due to some discrepancies. The procedure involved having the participants sit in a sound-attenuating chamber and a response device was put before each of the participants to keep track of the stimuli (Chwilla & Kolk, 2002). There was the pure list that comprised of 100 two-step pairs and the mixed record that was composed of 49 one-step pairs. The pure list always came before the mixed one to avoid any possible relationships between the two. To collect data for this study, the participants indicated the presence of a word by pressing a button with their dominant index finger, and for the other strings, which were not words, they pressed using the other index finger.

The results of the experiment show that two-step priming occurred in the double lexical decision task. In addition to that, it was also found out that the list composition did not determine the use of different strategies. It was also evident that two-step priming was larger than one-step and this form of decay in priming was consistent with spreading activation models and it was also typical for three-step priming (Chwilla & Kolk, 2002). However, it was noted that the two models that were to be evaluated could not account for the presence or absence of the list composition. As such, this implies that the hypothesis stated was not supported by the results of the experiment.

The authors are of the opinion that the current evidence that the above models can attribute for three-step priming is entirely new and is not in agreement with the traditional perception that mediated priming gives adequate support on the idea of spreading activation models. In my opinion, three-step priming is still a concept that has not been highly researched on, and as a result, the findings from the studies cannot be wholly relied upon because there is still need for further research (Chwilla & Kolk, 2002).

A Diffusion Model Analysis of the Effects of Aging in the Lexical-Decision Task

The argument is based on an article by Ratcliff, Thapar, Gomez and McKoon whose primary purpose of their study was to investigate the difference in lexical decision performance between older subjects (age 60-75) with younger, college-level students. They applied the diffusion model to compare lexical-decision performance for older participants from 60-75 years and college-age students. The lexical decision is one task that is important in this study to demonstrate that performance and the lexical-decision response would improve and increase with age.

Cognitive tasks proved that processing decelerate with age and that triggered this study. Another study by Allen, Madden, and Crozier (1991) found average RTs of 800 ms for older adults in comparison to 500 ms for younger participants. Longer RTs with lower frequency words are more substantial for older adults than younger ones. Also, a different study compared lexical-decision performance correctness between aphasic patients and normal control participants led to significant RT differences between the two groups. What has also been discovered in earlier studies is that older subjects have longer RTs, and that is possibly because of their slower indecision processing and because of additional conventional decisions that they have over young subject.

The hypothesis the authors’ state is very clear. If word frequency and type of non-word has an influence only on the rates of collection of evidence, then older participants would possibly have more substantial rates of accumulation of evidence that young participants, because of their countless years of knowledge with words, and so the hypothesis was examined by 2 experiments using a standard lexical-decision technique. There were many trials, and for each one, an individual letter string was introduced, and it was the participants’ choice to decide if what is presented a word or a non-word.

We were able to replicate the experiment in the three-step priming in lexical decision article by Dorothee J. Chwilla and Herman H. J. Kolk. Which was successful in by analyzing 37 female participants, aged 20-30, from Kuwait University and who were majoring in English Linguistics? The range of hours the participants spend communicating in English was between 1-10 hours per day. Bearing in mind that only one student converses in English for 10 hours per day, which is relatively good in consideration to our other classmates who fall in the medium of 5.5 hours, which is not sufficient for our level of education. When it came to the hours spent reading in the target language, the rage shifted to between 1 to 6 hours per day and 3.5 hours of reading being the medium. As per writing in English, the range is from 1 hour to 6 hours, which again is not much for our level. Between 4 hours to 15 hours are the total hours of English usage per day. The range of this mapped a critical example of the fundamental attributes and controls that were essential in researching with most of the students having to develop the critical contextual characteristics as underlined in the public school setting.

The experiment drew the participants from the public schools, and it was critical in creating a proper examination of the content and models that were used in the experiment. The CogLab experiment model was developed and creatively underlined through assessing and making concrete models and assessments. I tested it out on my laptop when I was at a café. We were presented with 50-60 white words and non-words that were all capitalized in letters that were displayed on a black background. We had to answer by pushing the symbol “/” for a word and /z/ if it was a non-word. The experiment used a three-word lexicon with the associated words RT, Unassociated words RT, and the words than non-words RT.

The results were that high-frequency words had shorter RTs and higher accuracy than low-frequency words, which had shorter RTs and higher frequency than very low-frequency words. Nonwords accuracy was higher, and RTs were shorter when the non-words were random letter strings, just like in experiment 2 than when they were pseudo words in experiment 1. Older participants correct RTs to words were longer by 150 to 300ms, and their right RTs to no words were longer by 150 to 250ms. As such, this answers the hypothesis that older participants are slower than the college students, and the accuracy of older subjects are better than younger subjects because the young college students’ error rates were twice as large as the ones made by older subjects. There was also a difference in the speed-accuracy relationship between young and older participants. The young participants’ responses for the least accurate were faster than their responses for the most accurate (Bentin, McCarthy & Wood, 2015). As for the older participants, the fastest was no less accurate than the slowest, excluding 2 participants. This shows that younger participants are likely to prefer speed accuracy than older participants. From the diffusion model, the results show that the nondecision component of RT was 80-100 ms longer for the older participants and they also adopted higher, more conservative decision criteria settings. It is implied that the older participants might not be able to increase their speed and respond quickly even when asked to do so (Bentin, McCarthy & Wood, 2015). Another finding is that drift rates that measure the quality of the match between a test string of letters, and lexical memory was about the same for older and younger participants, despite the years of experience that the older people have and substantial differences in accuracy.

From all the experiments, it was shown how the components of processing by the diffusion model work together to explain data. The quantile probability plots in the experiments of this study indicate that the model fits the data well, involving the right slant tails of the RT allocations and the distributional changed across experimental situations. The shape of the RT distributions in the experiments stayed almost constant. All in all, the model efficiently predicted mean RTs for correct and error responses, accuracy values, RT distributions, and how these variables change across experimental manipulations through distribution, not shape.

Finally, to conclude my review, I wish to declare that I liked the observations. I believe that the diffusion model is the best way to provide with correct and error RTs, RT distributions and accuracy values, which will show if accuracy is the same or different for older and younger participants. I liked how a comparison was made between the diffusion model and other diffusion models to show how they fit a lexical- decision experiment (Bentin, McCarthy & Wood, 2015). I also liked how subjects of similar RT performance, who made few errors were grouped to form super-subjects as this will assure the reliability and validity of data collection. The study didn’t rely only on the diffusion model but have used another procedure for examining RTs in ageing called Brinley plots.

The model was very beneficial in interpreting experimental outcomes for issues like ageing and speed of processing. However, when a comparison was made between the diffusion model and other diffusion models, not much was said about how similar or different they are from one another. I suggest more information about the various models to be explained so that whoever wants to replicate the study knows which model to go for. Error RTs in this study was hard to decode because of bias of movement toward one direction, which makes responses slower in the other direction. I propose a suitable device for more straightforward interpretation of error RTs.

Research questions

1. there have been differences in the easiness of using words and non -words by the students , how easy can a student use words appropriately compared to non-words at a given age bracket?
2. How do the student’s age and class level affect their usage of English words and non-words concerning the hours they practice reading and writing in the language?

From the results of the research, I choose to use a sample of nine participants which was small enough to allow me to accomplish the research within the limited time I had for the research, in addition, choosing only nine participants was enough as it accounted for almost half of class which had twenty students only. In the table above as an overall representation of the whole population that was subjected to the study. It is clear that the frequency of the students using words appropriately has a direct relationship with their age as a variable, while the hours they spend using the English language directly affects the frequency they use words subsequently before they use the non-words. The level of the study also concerns the use of associated words and unassociated words. Juniors appear to use non-words more subsequently, and an assumption can be made based on their exposure to the social media and hours spent writing and studying English.

With the research question, it is evident that the age and class level have a direct relation to the hours they spend reading and writing in English and also the social media platforms and hence affects their memory and use of words and unassociated words while undertaking the experiment. In perspective of lexical proceeding that is in the end settled on, applicable limitation emerges as an imperative factor in the handling of words and should be incorporated into any hypothesis of lexical handling (Bentin, McCarthy & Wood, 2015).

Discussion

From the results I managed to get from my research and in comparison with the works of McNamara and Altarriba (1988) and Chwilla, Kolk, and Mulder (2000) in regard to three- step priming in lexical decision, the impacts cannot be clarified as far as lexical co-event, which has been guaranteed to be a strategy for estimating separation in semantic memory better than free affiliation (Bentin, McCarthy & Wood, 2015). Three-step priming decisions were observed regardless of the utilization of rigorous criteria in view of the free-affiliation strategy or the lexical co-event technique.

This outcome underpins the view that the three-step impact was caused by spreading enactment and has all the earmarks of being conflicting with the recommendation that familiarity with the words may represent three-advance preparing. The implication of this field is that students’ use of language and associating their words are as a result of how much time they spend engaging in the language. To summarize the entire study, from the results of the research representing the overall population, it is evident that the period in hours that students engage themselves and familiarize in the language affects their spoken language with regards to their age and level of study. The nature of school they attend however does not affect how they use the language.

References

Bentin, S., McCarthy, G., & Wood, C. C. (2015). Event-related potentials, lexical             decision and semantic priming. Electroencephalography and clinical           Neurophysiology, 60(4), 343-355.

Chwilla, D. J., & Kolk, H. H. (2002). Three-step priming in lexical decision. Memory         & Cognition, 30(2), 217-225.