The results of the present study revealed worse performance by individuals with no formal education than low-educated subjects on several oral language tasks from the MTL-BR Battery. This study corroborated the findings of previous reports in the literature (Radanovic, Mansur, and Scaff, 2004; Soares & Ortiz, 2009) showing that education influences the performance of healthy individuals on language tasks, namely: oral comprehension, oral narrative discourse, word repetition, phonological/orthographic verbal fluency, number dictation, number reading, and written numerical calculations (Table 1). Individuals with no formal education had worse performance than low-educated subjects.
In a previous study, Akashi and Ortiz (2018) investigated the influence of low education on the language tasks assessed by the MTL-BR Battery and highlighted the importance of developing more studies with larger populations, including more individuals with 1 and 2 years of formal education and with no formal education, since there are still no normative data available for these populations for the MTL-BR Battery use. Their results lend support to the hypothesis that even a few years of education affects language performance and hence impacts scores on the different MTL-BR subtests. In this study, statistically significant differences in performance by low-educated individuals relative to individuals with no formal education were observed and a minimum level of education appears to change cognitive performance. These findings will be discussed below.
There was a difference between the performance of NFE and LE groups in the auditory comprehension subtest. This task measures the ability to identify images that represent words and phrases from auditory input. However, the qualitative analysis showed better performance on oral comprehension of words than of sentences for both groups, with several possible explanations for this disparity. Pictures depicting actions require a larger number of visual inferences than single items, while the core components for understanding an image are inherently more detailed, which in turn requires correct visual perception of both agent and object (Mansur et al., 2005). In addition, greater demand is placed on working memory to process sentences with non-canonic structure, such as those with passive structures and center-embedded clauses, since the phrase must be first stored, then organized and syntactically decoded, for final comprehension of the information and selection of the correct drawing (Eom & Sung, 2016; Ortiz & Bertolucci, 2005). Previous studies have found differences in oral comprehension of sentences when groups with 1–4 years of formal education and 5–8 years of formal education were compared (Akashi & Ortiz, 2018) and even in comparison of 5–8 years of education to 8–12 years of education (Pagliarin et al., 2014) and, in this case, it can also be influenced by reading and writing habits (Pagliarin et al., 2015). Finally, the oral comprehension of text requires comprehension, retention and retrieval of the information presented in the text, recruiting working memory and components of executive functions. However, working memory and metacognitive operations are mediated and boosted by formal education (Ardila, Rosselli, and Rosas, 1989; Jou & Sperb, 2006), possibly explaining the lower performance by individuals with no formal education compared to low-educated individuals on this task.
Regarding to the oral language tasks investigated in this study, the participants presented low performance on oral narrative discourse task. The narrative comprises a description of a series of events and actions. The manner in which individuals explain the actions of the characters resembles the way in which they construe actions of people in everyday life. Therefore, for this process to function effectively, inferences must be created (Bower & Morrow, 1990). These inferences can be explained as mental representation formed through interaction between explicit linguistic information and world knowledge held by the individual (Alonso, 2004; Zanichelli, Fonseca, and Ortiz, 2020). These inferences aid the construction and comprehension of discourse (Alonso, 2004) and depend on the world knowledge held by the individual and might explain why, during this test, the discourse produced tended to be based on the individual´s personal experiences of the world/everyday life when the meaning of the image could not be grasped, resulting in stories unrelated to the theme presented. The individuals may have encountered difficulty understanding the context of the target situation and/or problems accessing their scripts (a bank robbery—in the case of the MTL-BR) on world knowledge or use them properly to make pragmatic judgments correctly (Hirst, LeDoux, and Stein, 1984). It is likely that, due to a break down in the inferential process, the individuals produced descriptive discourse. The failed integration of the elements present in the stimulus must have led the subjects to describe each object in the drawing in detail, without correlating them to form a story. The main components of a target figure are directly related to the generating of inferences (Ribeiro & Radanovic, 2014). Thus, if the individuals had fewer information units regarding the scene from the outset, they likely made fewer visual inferences, explaining the poorer narrative discourse produced. In fact, successful discourse requires the combination of information units, as propositions, in a coherent way to convey a significant message (Wright, 2011).
On the repetition task, the individuals with no formal education had major difficulty and presented lexicalization errors, suggesting the use of the lexical route when the phonological route was needed. The phonological route starts, naturally, by a phonological analysis of the auditory input. A phonological input buffer permits the storage of phonological information (segmented and correctly sequenced) for a short period (Reis & Castro-Caldas, 1997). The repetition of non-words demands comparisons or the detection of the specific phonemic characteristics of words. It seems that it is precisely this analysis that is problematic in illiterate individuals, who demonstrate difficulties in certain tasks that required phonological awareness (Reis & Castro-Caldas, 1997). In fact, the MTL-BR Battery has pseudo-words that depends on the phonological route responsible for phonemic coding typically underdeveloped in illiterate or low-educated subjects (Petersson, Reis, Askelöf, Castro-Caldas, and Ingvar, 2000).
The individuals with no formal education also had worse results on Arabic number dictation and Arabic number reading, as well as on mental and written mathematics calculations. First, the option to investigate the performance of individuals with no formal education on these tasks is because numbers are present and mathematical calculation exists in many everyday activities. Calculation ability under normal circumstances requires not only the comprehension of numerical concepts, but also that of conceptual abilities and other cognitive skills, so it is impossible predict the exact impact of daily activities on number learning and processing. The difficulties on Arabic number reading and Arabic number dictation were more marked for numbers containing hundreds and thousands. These findings corroborate previously results (De Luccia & Ortiz, 2009) showing that low education impacted performance on some mathematics tasks, such as orthographic transcoding of numbers. On mental mathematical calculations, the subjects had no problems for addition or subtraction, but all encountered difficulties with multiplication and division. This is particularly true for carrying out multiplications, which need knowledge of the times table, the most commonly used approach for teaching multiplication in Brazil. It is worth mentioning that individuals with no formal education performed mental calculation as well as with low-educated individuals (Table 2). This pattern probably occurred because simple addition and subtraction are more commonly used in everyday situations than other operations requiring more formal learning. This finding is important in as far as it supports the notion that, although analyzing the effect of education on individual performance during neuropsychological tests is paramount, the influence of social environment should also be investigated. This environment dictates whether the individual received stimuli to develop certain abilities or otherwise, further contributing to cognitive performance. For written mathematical calculations, individuals with no formal education were unable to solve, irrespective of mathematical operation involved (addition, subtraction, division, or multiplication), possibly because at this part of the task, the mathematical operations are more complex and then, probably more dependent of learning obtained through a formal education. Indeed, level of familiarity with carrying out arithmetic increases with years of formal education (De Luccia & Ortiz, 2009).
On the semantic verbal fluency task, participants had difficulty producing words from the animal’s category and, on numerous occasions, participants were in doubt over whether a given word belonged to the category. This difficulty might be explained by the fact that formal education facilitates the organizing of semantic subgroups and categories (Ratcliff, Ganguli, Chandra, Sharma, Belle, Seaberg, and Pandav, 1998). Nevertheless, no statistically significant difference between no formal education and low-educated groups was evident. Considering this task, comparing the data from this study with data from healthy individuals with 5–8 years of education, differences were found and they are possibly owing to sociocultural influence and learning through lifespan. Animals are very familiar category. As outlined earlier, although education is a determinant of cognitive performance, it is not the only variable to consider. The role of stimulation and probable influence of the sociocultural environment on cognitive development should also be taken into account. These aspects, however, are difficult to measure objectively.
For phonological verbal fluency, the subjects exhibited great difficulty producing words, most likely explained by the connection between development of phonological abilities and formal education. Most studies investigating the impact of literacy on oral language processing have shown that literacy provides phonological awareness skills in the processing of oral language and the ability to segment speech into phonemic units is dependent on literacy (Tsegaye, De Bleser, and Iribarren, 2011; Ratcliff, Ganguli, Chandra, Sharma, Belle, Seaberg, and Pandav, 1998). Our results suggest the existence of differences in phonological processing even between individuals with no formal and low formal education, as was demonstrated by Ardila, Ostrosky, and Mendoza (2000) and Colaço, Mineiro, Leal, and Castro-Caldas (2010).
Finally, on the other hand, Table 2 shows tasks from MTL-BR Battery that no differences were found between individuals with no formal education and low-educated ones. Since MTL-BR Battery was published, several studies investigated scores in large populations, considering the variables age, years of schooling (Akashi & Ortiz, 2018; Pagliarin et al., 2014) and other sociocultural variables such as reading and writing habits (Pagliarin et al., 2015). Taken together the data from these previous studies and the present study, it can be observed that the tasks structure interview, automatic speech (content), nonverbal praxis, object manipulation and body part recognition seem not be influenced by years of schooling. The structure interview is a task that is based on a series of questions, most of them about people’s daily lives. It is a conversational task, and probably the questions can be easily understood and answered by people, regardless of sociodemographic variables. In the case of the object manipulation task, auditory, proprioceptive and visual processing are involved in its execution. The familiarity of the objects presented and the tangible effect they evoke may have also facilitated execution of the task. These factors likely promoted the similar performance in carrying out the task (Medeiros & Ortiz, 2021). In addition, results of a previous study (Akashi & Ortiz, 2018) revealed the presence of a ceiling effect on this task among healthy individuals with low educational level. In turn, the absence of differences between groups in automatic speech and non-verbal praxis tasks were expected. Automatic series are considered the linguistically simpler tasks and non-verbal praxis only requires individuals to perform movements involving the tongue and lips. So, even considering more ecological or informal methods to assess PWA with no formal or low education, these tasks can be used and probably can be helpful to identify language and speech disorders and follow-up of these patients.
This study has clinical applicability because the MTL–BR is the only language assessment battery for adults available in Brazil, rendering it especially important to determine the effects of education on task performance. Although a pilot study, the data gathered can be a guide during the application of the MTL-BR Battery in PWA with no formal education.
This was a pilot study investigating the influence of no formal education on performance on language tasks measured by the instrument. However, studies involving larger populations and studies that control possible influence of the socioeconomic status are needed to confirm these findings with individuals with no formal education on the MTL-BR Battery. The significant differences in performance between no formal education and low-educated groups on the language tasks assessed strongly suggest that populations with little formal education should be studied in greater depth, perhaps according to each year of education as opposed to the broader education bands (1–4 years) used in most studies.