In comparison to other studies on memory functioning in dyslexia which referred to the FTT (e.g., Blau, 2013; Brainerd et al., 2006; Miles et al., 2006; Voss, 2013), the present study shows a model-based examination of differences in memory processes between students with and without dyslexia. As mentioned earlier, the usage of multinomial models allows for an analysis of the basic latent processes that are assumed in the FTT. The results of previous research conducted on this topic only gives general information on verbatim and gist trace functioning, without any distinctions between specific processes. Because of this, the conclusions formed on the basis of these studies could only show the global verbatim and gist memory functioning pattern. The results of the presented study give an insight into the hypothetical latent processes underlying memory performance. Differences were found only in some of the manifestations of verbatim and gist trace retrieval while, at the same time, differences in other processes connected therewith were not observed. This means that the impairments (and/or enchantments) of memory processes are not global in terms of verbatim and gist—which was suggested by the findings of other researchers (Blau, 2013; Brainerd et al., 2006; Miles et al., 2006; Voss, 2013).
Impairments of verbatim trace retrieval are located in the same process for both the semantic and the orthographical conditions. Deficits were found in the case of the recollection rejection process, which is responsible for the rejection of a stimulus which is only related to the target presented at study. In everyday situations, this process allows to differentiate between similar objects. This ability seems crucial in reference to the deficits and difficulties affecting people with developmental dyslexia (see American Psychiatric Association, 2013; World Health Organization, 1992). Problems with literacy like, for example, the erroneous reading of a word, can be explained in terms of the FTT. As mentioned earlier, Brainerd and Reyna’s theory is closely connected to psycholinguistic theories (Reyna, 2012); therefore, our findings could easily be connected with language learning deficits. The reading of words, on its basic level, can be connected with the verbatim trace. The developed reading skill (e.g., Indrisano & Chall, 1995)—in which reading does not involve the separate analysis of every character of a word—is based on the recognition of the whole physical structure of the word while reading it. Also, in the case of the early stages of reading skill development—in the grapheme-to-phoneme conversion process—verbatim memory is probably necessary to remember which character sign refers to a particular sound. In both cases, the similarity between given a word/characters and other objects in memory could cause errors in the reading process. Therefore, the recollection rejection process seems to play crucial role in avoiding errors caused by the similarity of one word or character to another. A deficit of differentiation between objects could not only lead to reading skill but also to writing skill errors (also observed in dyslexia); for example, one can use similar characters or character sequences in the place of the correct one and make a spelling mistake (e.g., in the Polish language, the characters “ż” and “rz” are connected with the same phoneme but cannot be used interchangeably).
Recollection rejection process, or more wide verbatim trace, plays important role not only in processing orthographically similar but also semantically similar words. Synonyms are words close but not exactly the same in meaning, but different in phonetic and graphic construction (e.g., large and big). Verbatim trace of a given word plays crucial role in distinguishing between semantically similar words because it gives additional information, that are different than nuance differences between meaning of given words. Thanks to retrieval of verbatim trace memory of gist information becoming more precise and clear, which results in less number of errors based on semantically similarities.
Another hypothesis could also explain the deficits observed in the performance of reading and writing of persons with dyslexia—i.e., grapheme-to-phoneme conversion (Tholen, Weidner, Grande, Amunts, & Heim, 2011); however, this approach fails to explain the shortcomings of the recollection rejection process. In turn, the findings concerning memory functioning in dyslexia can be used to explain other such process impairments described in literature. There is also the possibility that the deficits of these different processes coexist in dyslexia and their joint impact causes the observed deficits. The memory functioning differences on a gist trace level are not as clear as the verbatim trace findings. Similarly, only two of all the varieties of the hypothetical memory processes that are based on gist differed between experimental and control groups. The probability of gist trace retrieval when the stimuli is a target in the semantic condition (Gt
s
) was higher in the control group; on the other hand, the probability of gist trace retrieval when the stimuli is related to the target in the semantic condition (Gr
s
) is lower in the control group than in the dyslexia group. The differences in the functioning of memory of this trace are not as intuitive as those in verbatim trace memory. However, the assumptions of the FTT described by the multinomial model show possible explanations of the obtained results.
A close look at impaired and enhanced processes allows the pattern, which can be connected with compensatory mechanisms responsible for overcoming problems that people with dyslexia deal with, to be identified. Verbatim trace process differences are easily to connect with the symptoms of dyslexia, i.e., literacy skill deficits. The gist trace memory functioning pattern is more complex. It cannot be connected with directly, only through verbatim processes. The observed impairment of verbatim trace retrieval suggests that people with dyslexia have problems with correctly answering questions when differentiation between target and similar information is required. Improved gist retrieval, as measured by the Gr
s
parameter, allows a correct answer to be given and, thus, enables an adolescent with dyslexia to overcome his or her difficulties. Unfortunately, improved gist trace memory also affects the probability of a memory mistake because gist trace could either lead to right or wrong answers—the correct recognition of the target—but also to the false recognition of a related item (Brainerd et al., 2003; Stahl & Klauer, 2008, 2009). On the other hand, poor gist retrieval when the target is presented (in the semantic condition) leads to a lower probability of both the correct answer and the false recognition of a related distracter. Therefore, this second gist trace difference (in the Gt
s
parameter) may be a response to the Gr
s
parameter difference. The lower probability of gist retrieval in a condition that is not influenced by verbatim trace retrieval deficits decreases the overall probability of a correct response but also counteracts the negative effects of gist trace increase in the Gr
s
parameter, thus enhancing memory differentiation.
From the other hand, there is possibility that obtain sample constitutes a large number of students with deep dyslexia. People with this type of disorder display not only deficits in accuracy and fluency but also semantic errors (e.g., Buchanan, McEwen, Westbury, & Libben, 2003). If this situation occurred, it could explain improvement in gist trace retrieval when related stimulus is presented in semantic condition—this difference could lead to more errors with basis in gist trace, and connected with impairments of recollection rejection it fits deficits pattern found in deep dyslexia. However, second difference—impairment of gist trace retrieval when target stimulus is presented in semantic condition—leads to lover probability of gist-based memory error; thus, observed differences are not precisely match to deep dyslexia symptoms.
The FTT gives also a possible answer to the inconsistency in the findings about reading comprehension performance in dyslexia. On the one hand, some theoretical conceptions and empirical findings have assumed that process of reading comprehension in people with dyslexia is impaired in comparison to typically learning individuals (e.g., Fletcher et al., 2002; Wong, Ho, Au, McBride, Ng, Yip, & Lam, 2017). On the other hand, other studies (e.g., Miles et al., 2006 and the presented study), conducted under the FTT, suggest that gist trace memory (connected with comprehension; see Weekes, Hamilton, Oakhill, & Holliday, 2008) could be the same or even better in dyslectic than in typically learning individuals. However, this inconsistency can stem from the age differences of studied groups. Research based on FTT shows that there is the swap of memory reliance from verbatim trace to gist trace (Reyna, 2012) during development. In fact, young children could rely mostly or even entirely on verbatim processing in the comprehension of a read text; thus, reading comprehension observed in studies on young children (Wong et al., 2017) could not be related with the performance of gist trace memory.
In the context of education, obtained result could suggest possible approaches in methods of teaching of adolescence with dyslexia. Firstly, teachers should remember about deficits in recollection rejection process (thus, impairments of verbatim memory) that affect pupils with dyslexia. This kind of impairment could lead to errors not only in reading but also spelling and even memory of other type, precise information—if deficits are not narrowed to verbal memory. Because of that, teachers should based their evaluations of pupils’ works more on the gist information, and less on memory of exact information (like date in history, or remembering of names and text in the language classes). In the same time, teachers should give to their pupils with dyslexia some additional exercises that focus their attention on mistakes their made and help them to remember correct forms, answers, etc. Even if deficits will affect only reading, this approach still can bring good results. From the one hand, students with dyslexia will not feel worst then other student; thus, their situation will less affect their self-esteem. But, on the other hand, they will still have opportunity to work on their mistakes to overcome their difficulties. Secondly, knowledge about which particular process are impaired should lead to more focus forms of therapy and exercises that aim to improve this specific disabilities. Results of Moore and Lampinen (2016) research suggest that people are aware of using recollection rejection in their attempt to remember something. Therefore, it is possible that someone could train this memory process to improve its use in everyday situations. However, this approach needs further investigation.
The presented studies offer new opportunities based on the FTT memory model for the examination of the connections existing between developmental dyslexia in adolescents and their memory functioning. The application of the multinomial model gave an insight into the more elementary memory processes assumed in the used theory, which allows for a deeper and more precise examination of memory functioning in dyslexia than in other studies using standard analysis methods. Nevertheless, more studies based on FTT and multinomial modeling have to be conducted to comprehensively analyze the connections between developmental dyslexia and the functioning of verbatim and gist memory. The present research only covers one age group—adolescents. There is a possibility, supported by theoretical and empirical evidence, that the pattern of memory functioning in people with dyslexia in different age groups could be different. This is precisely why future studies should compare memory functioning not only between participants with developmental dyslexia and their peers but also between participants in different age groups.