International Journal of Innovation and Economic Development
Volume 5, Issue 6, February 2020, Pages 43-54
Concept Mapping for Noun Identification in CLIL Textbooks by Primary Education Students
1 José Luis Gómez, 2 Enrique Arias, 3Juan Lirio
1 Faculty of Educación, University of Castilla-La Mancha, Albacete, Spain
2 3 Faculty of Social Sciences, University of Castilla-La Mancha, Toledo, Spain
Abstract: This study aims at observing if initial training on concept mapping improves pupils’ skills on noun category identification from foreign language (L2) texts. To observe the null hypothesis, learners receive short-term training to manage the basics of concept mapping. After the intervention, in a classroom context of reading and writing assignments in the English language, Spanish primary students identify and circle substantives in texts from their CLIL textbooks. The sample comprises sixty fifth-grade Spanish pupils age-ranged 10 to 11. Though the school has an English-Spanish bilingual program, the students are not complete bilinguals. In the sample, participants belong to both male and female genders non randomly assigned. Concerning the identification of singular and collective nouns from texts, including plural irregular nouns, both treated and control sample groups displayed similar results. Although the instructed group was skilled in detecting and categorizing hypernyms and first hyponym categories of nouns, they did not precisely discriminate adjectives, confusing them with nouns. Dissimilarly, the untrained control group not only mistakenly identified adjectives as nouns but also other grammatical categories as verbs, conjunctions, pronouns, and adverbs. The scarce disclosed precision by the non-trained group was also perceptible in the production of the concept maps. When they had to make the concept mapping tasks, control participants tended to make flow charts, mind maps, and tree diagrams instead. Contrarily, the trained group accomplished the activities satisfactorily. Yet the outcome of this study reveals that a little training on concept mapping leads to disclosing slight achievement in discriminating information from L2 texts in students. Hence, statistical analysis reveals that concept mapping helps students differentiate lexical and grammatical categories and to synthesize the information.
Keywords: CLIL, SLA, Concept mapping, Bilingualism, Primary education
To some Spanish primary students in bilingual schools, content and language integrated learning (CLIL) textbooks may include specific and unknown new concepts or information that they must learn. To facilitate learning, teachers, through a didactic transposition process and particular teaching techniques, disclose the essential information that bilingual pupils need to know. In some other cases, teachers might even show students the flow of information in L2 texts, and how hypernyms enclose hyponyms within paragraphs and how words are linked and organized within texts.
For organizing the information, concept mapping constitutes a valuable didactic support for teachers when giving visual or verbal explanations. However, because of a lack of time or classroom resources, it is not always feasible to teach pupils concept mapping techniques for the subsequent management of the information by themselves. To Mintzes, Wandersee, Novak (1998), skills as the identification, selection, and extraction of ideas from texts are necessary for the understanding of more sophisticated and complex material. Hence, this research aims at approaching concept mapping to Spanish primary students from a bilingual school.
Considering the previous information, the goal of this study is to monitor if students trained briefly in the concept map usage can discern specific information from English texts in a more efficient way than their non-trained peers. The reason we consider discrimination of information at this age as stage-important is that it might affect the present and future students’ knowledge and final marks when learning curricular content in L2. The tool we propose to facilitate this is the use of concept maps. Also, we believe that both learning curricular content in L2 and managing concept maps demand high order thinking skills from primary students.
The stress of facing new content and self-learning techniques was observed in this study, causing more stress in the experimental group than in the control group because of the extra, unknown, demanding homework needed for the use and familiarization of concept maps. For this reason, when teaching the use of concept maps to primary students, we also suggest a step-by-step concept mapping instruction from the very beginning, as well as giving pupils time to develop their skills in the use of this helpful tool. Therefore, the hypothesis inferred is that the trained experimental group will discern nouns from English texts more efficiently than non-trained students belonging to the control group. As a result, they will summarize more proficiently the information to be learned.
2. Literature Review
When acquiring knowledge in an L2, it is generally beneficial for students to summarize the curricular content managed, especially the specific information to be recalled when sitting an exam. Summarization includes using mind processes, language interpretation, and narration (Kay, Roberts, Samuels, Wotherspoon, 2009). In summarizing content, summarists have to know how briefly to restate the essence of a text in a new and personal manner (Figure 1). In order to support learning and organize information, most CLIL teachers use multi-media and visual organizers (Bentley, 2013).
Figure 1. Concept map made by a two-month-trained student.
Regarding summarization, including the above mentioned culling out skills to be taught to pupils as self-regulation, specific teaching techniques to summarize texts for the subsequent study are not commonly taught explicitly in the average classroom. However, research shows that summarizing information facilitates learner comprehension and long-term retention of information (Wormeli, 2005). Our premise is that gradation, when summarizing from lexical categories to summaries, employing concept mapping promotes meaningful learning (Figure 2).
Figure 2. Concept map and summary made by a six-month-trained student.
Text comprehension in a language includes the knowledge of basic grammatical and lexical categories. Furthermore, this grammar affects the way reality is perceived. Each personal view of the world is dependent on the language structure of the speaker allied to a language (Seuren, 2013). This grammatical-individual interpretation of reality implies biological matters as well, for example, brain development. In some cases, both grammatical awareness and biological development represent barriers to foreigners when interpreting and summarizing curricular content in a foreign language.
The lack of grammatical awareness and text comprehension occurs especially in early ages, when teachers at bilingual schools teach subjects other different from the English language, in English, to non-native speakers. There are also external barriers for pupils to deal with, such as relatives who cannot correctly help these pupils to finish their homework because they are not bilingual.
Spanish primary students’ misinterpretation of English texts diminishes their comprehension and long-term retention of curricular knowledge, even if they summarize the information. For that reason, the use of concept maps is recommended in education from an early age (Novak, 2010). For starting teaching the usage of concept maps to non-English speakers, it is reasonable to initiate the instruction of pupils to the English-essential grammatical and lexical categories, especially nouns. In previous research where participants had to recall isolated words, the scores were higher for noun recognition than for verb keyword groups.
An explanation for the earlier example is as follows: “A plausible cause of this effect of a grammatical class is that the referents of verbs are harder to imagine than those concrete nouns (. . . .) more effective keywords than verbs” (de Groot, 2011, p. 97). Although both groups identified most of the nouns within the four texts, in this research, students exhibited different results when confusing other grammatical categories with nouns. For example, the experimental group identified adjectives as nouns, whereas the control group identified verbs, pronouns, prepositions, conjunctions, and adverbs (mainly grammatical categories other than adjectives), as nouns.
To identify lexical categories, precisely nouns within texts, pupils focus attention on two linguistic aspects: grammatical structure and translation of words to recognize their meanings and semantic fields. Structurally, students segment texts by scrutinizing the syntax to identify words according to their position within the sentence. Noun recognition is not an easy task for primary pupils. They identify, underline, highlight, or extract each word categorizing the grammatical structure and word functions.
Semantically, children associate words with their meanings using looking for their denotative significances in bilingual or monolingual dictionaries. They also request a translation from peers and teachers. In some circumstances, “When children are beginning to discover written [foreign]language, helpful adults often act as mediators by saying what the printed words are, leaving to the child the more complex task of [grammatically]discovering how to distinguish one word from another” (Smith, 2004, p. 132).
In consequence, pupils that commonly use concept maps to deal with information are supposed to identify word categories within natural science texts effortlessly. This instruction makes them aware that word functions are different. They are conscious that nouns are dissimilar to the rest of the grammatical words because they are words enclosed within boxes in the concept map frame. Hence, learners understand that, when those concepts are in order, some words are more inclusive than others (Figure 3).
Figure 3. Three-month-trained student’s creation when asked to make a concept map.
Instructed pupils do also discern that those concepts display connections through linkers, which contain different word categories other than concepts (Novak, 1988). Undoubtedly, the more prior knowledge pupils have about the curricular content, the easier it is for them to identify nouns and other lexical categories within a text, thus easier to acquire more significant learning (Ausubel, 1968). Nevertheless, to learn new curricular content in a foreign language can be confusing for students whose English level is below the lexical-grammatical content in their class books unless they know the use of concept maps previously.
A more explicit example is below, where some trained and non-trained students in the use of concept maps had to make a concept map. The task consisted of making a concept map from an approximately 158-word-lengthen text, including the title and headings. Both groups performed this task at the same time, and none of them, in theory, had had previous knowledge about the curricular content. While doing the task, they were not allowed to ask teachers or classmates for help or use a dictionary.
As a result of the previously mentioned activity, the three-month trained student in the use of concept maps appropriately encloses concepts, but still identifies some adjectives as nouns. A reasonable explanation for this could be that the headings they had been using to make concept maps from their class books (large letters and green or black bold print) were up of two or more words, usually composed of an adjective and a noun (Living Things) but sometimes included other grammatical categories of words (Parts of Plants). Similarly, the non-trained student chooses nouns and adjectives as hypernyms, but chooses a different visual organizer and adds new words in it (Figure 4).
Figure 4. Non-trained student’s creation when asked to make a concept map.
3. Research Methodology
This experiment consists of quantitative research. Considering noun category identification within written English texts, we aim at determining concept mapping trained students’ comprehension and abstraction. We have designed a post-test case study in which formative teaching applies to one of the analyzed groups. The analysis is descriptive, to provide the analyzed students’ profile. This contingency analysis, or cross-tabs, allows us to establish a relational level between variables. Finally, the average comparative analysis will show the difference existing between the two groups, the experimental and the control ones.
For counting and evaluation of the detected nouns and non-nouns, we have used three equal intervals, based on the selected nouns’ rank according to each of the fulfilled activities. We have gradually named these intervals so that correct answers and errors for the ulterior identification. Observe that error intervals are the opposite of correct answer interval, the fewer the errors (developing) the better the result.
- Correct intervals are identified as follows:
- Excellent = 3/3
- Good = 2/3
- Developing = 1/3
- Error intervals are identified as follows:
- Excellent = 3/3 errors
- Good = 2/3 errors
- Developing = 1/3 errors
Two equivalent groups of subjects–experimental (n = 30) and control (n = 30)–differently treated. Sample groups consisted of the fifth-grade students, all Spanish-language speakers, aging between 10-11 (M age = 10.5), selected from a bilingual school located in Castilla-La Mancha, Spain. Experimental (male = 19; female = 11) and control (male = 20; female = 10) groups were randomly assigned. Nevertheless, individuals belonging to each group were not randomly assigned. In the table below (Figure 5) percentages of representation (males and females) belonging to each group, and collectively, can be seen.
Figure 5. Experimental and control percentage representation.
Participants were unmodified, naturally existing groups: two fifth grade classes. Groups did not significantly differ in any incidental characteristic. However, a special-need pupil coexisted in the experimental group during the experiment. Tasks for this participant were in the Spanish language. Another particularity is that a slightly higher subject mark average from the previous school year in the control group (0.4) compared to the experimental group. 100% of the group members participated in the experiment.
3.2 Materials and procedure
The four texts used in the experiment were from the pupils’ Natural Science book (Riach & Bacon, 2014), whose curricular content comes from the Primary Education 54/2014 Decree. Texts are from the book mentioned above because it is the textbook that both groups were using during the 2015-2016 school year. Curricular content belongs to Unit 5, “Ecosystems.” This unit was implemented at the same time when the experiment took place. The unit of knowledge encompasses pages 68 to 73 of the book.
The text format, adapted for the experiment, was the same as the textbook format, but drawings and other noun-identification facilitators were omitted from it. Highlighted and bold print within the task texts was also deliberately excluded. Only titles, paragraph headings, and some hypernyms appeared in bold within the task texts for reference purposes. All the texts are in English; however, to facilitate students’ understanding, the instructions are in Spanish were included.
Assignments consisted of four readings with two tasks to perform in each, eight tasks total. Tasks 1 (4) varied from text to text (qualitative); tasks 2 (4) were common to all texts (quantitative). A total of eight tasks or steps were carried out as normal classroom activities, lasting approximately three hours total (T average = 22 minutes and 5 seconds per task). The sole purpose of the four task ones (T1) was in order for students to establish the first contact with the curricular content. Tasks develop as follow:
In the first assignment (T1), students have to summarize text A into a paragraph without text translation or explanation. Next, translation and explanation of the text took place with their teacher. Finally, completion of task two (T2), using the same text A, consisted of circling all the nouns within the reading.
In the second assignment, text B, students have to condense the text for a study purpose without text translation or explanation, as they usually do when summarizing information (T1). Next, they had to describe the process or steps they followed in order to do this task. Finally, translation and explanation of the text took place with their teacher. Then they completed task two (T2) by circling all the nouns from the reading.
In the third assignment, pupils have to extract the nouns from text C without any translation, explanation, or use of dictionaries (T1). Next, translation and explanation of the same text took place with their teacher. Finally, they completed task two (T2) by circling all the nouns from the same reading C.
In the fourth assignment, without text translation or explanation, students were asked to create a concept map of text D (T1). Next, translation and explanation of text D took place with their teacher. Finally, they completed task two (T2) by circling all the nouns from the same text D.
Assignments from texts A, B, C, D consisted of two tasks per text. This process was always routine (tasks one [T1] consisted of pupils having a first contact with the texts, then translated into Spanish, and finally, tasks two [T2], consisting of circling all the nouns from the texts).
Regarding materials, no specialized resources were necessary. Students used blank sheets of paper and texts A, B, C, D with the instructions in Spanish, to fulfill each session’s tasks. Four different texts for the four assignments (A, B, C, D), whose nouns were to be circled by the students during each session–quantitative tasks two (T2) named A2, B2, C2, and D2. Also, the additional piece of paper to complete qualitative tasks one–named A1, B1, C1, and D1.
Students had to have a pen, pencil, eraser, and colored pencils on their desks before starting the tasks. To finish qualitative tasks, students belonging to the experimental group asked for extra blank sheets of paper to complete some of the tasks. The vast majority of components in this group had difficulties completing the task within the space designated to do tasks B1 (condensing the text for a study purpose in the way they usually did) and D1 (making a concept map), in comparison to the control group.
4. Data Analysis and Interpretation
As shown in the next bar chart, in both selected groups, male representation was higher (65% male, 35% female from  total). All participants are in the same grade and are of similar age (10 to 11 years old [M age = 10.5]). There is almost no difference in their extracurricular training classes, with only a slightly higher percentage for those belonging to the control group, except for extracurricular English classes (see Figure 6). The percentage results referring to this aspect are as follow:
The experimental group (some students might attend both classes, English and other subjects, simultaneously)
- 33 % of the students do not attend extracurricular classes
- 66 % of the students attend extracurricular English classes
- 20 % of the students attend extracurricular non-English subjects (maths, Spanish)
The control group (some students might attend both classes, English and other subjects, simultaneously)
- 63 % of the students do not attend extracurricular classes
- 36 % of the students attend extracurricular English classes
- 23 % of the students attend extracurricular non-English subjects (maths, Spanish)
Figure 6. Students per group attending or not to private English, and other subject lessons.
Regarding the results of the four tests, the nouns correctly identified in each text show us that higher percentages locate within the superior interval entitled “excellent” (see Table 1). There was also a more significant number of undetected nouns, and thus errors, the majority of both groups tend to be within the higher interval as well, except for the A2 test where they are indeed within the lowest interval. This fact reflects the lower degree of complexity and less number of words, and thus nouns, in the A text.
Table 1. Percentages of correct answers and number of errors per group and task.
From the contingency analysis, no significant relationship exists between the gender variable and the correct vs. error variable when identifying nouns, with the signification indicators in the chi-square test above 0.05. These variables in the four tests run have no significant relation to the extracurricular English class variable either, and therefore do not discriminate. The levels of signification also rise above 0.05. The same occurs with the variable group (experimental or control). There is no table with an existing significant relation, and therefore, having been trained to use concept maps during three months does not imply having better results in noun identification in either of the texts.
Finally, in the t-test for independent samples, the results give us average results above 2.5. This factor indicates that noun identification would be at the highest levels in both groups (excellent). The averages of the non-identification of nouns would oscillate in the middle levels. The same occurs in both groups. Analyzing the differences in the averages, we can observe that the difference is only significant in the A2 test, with a difference of -0.333 (Table 2). Here, a distance is established between the groups, favoring the experimental group over the control one. As for the rest of the texts (B2, C2, D2), the differences are not significant. It may be that there is no difference between the groups. Therefore, we are working with considerably homogenous groups concerning the number of nouns identified.
Table 2. Levene test for equality of variances.
5. Conclusion and Recommendations
Some relevant conclusions are from the obtained results. Firstly, we are using two almost entirely homogenous groups in regards to their capability of identifying nouns within the texts, all of them displaying acceptable results. This particularity alone proves real success through training/instruction of concept maps to the experimental group. This last group started the academic year with a difference of 0.4 average subject marks lower, compared to the control group. Within the A, B, C, and D texts, these trained students mostly made the mistakes by circling adjectives as nouns, compared to the control group. To make use of concept maps more easier, these pupils were allowed to include within the boxes whole propositions (mainly headings formed by an adjective and a noun).
Due to these factors, the statistical figure shows in the t-test that there are hardly any differences between the two groups, which implies the acquisition of comprehensive and abstraction skills by the experimental group. This characteristic means that the experimental group, which had demonstrated a lower academic level before the instruction, was able to obtain the same results as the group having a higher academic level. The results also conclude that neither the gender nor the extracurricular English classes would be variables that discriminate, regarding the number of nouns identified in the tests. Therefore, the level of comprehension is not determined by these variables, confirming that gender does not determine the levels of cognition.
At the methodological level, the discussion could arise about if the division into three intervals is adequate or if to obtain more accurate information; it would be interesting to increase these intervals to six. Likewise, there is a doubt about whether the execution of the noun discrimination activities within the texts should be before or after text translations and conceptual explanations. Another methodological discussion emerges from whether it is appropriate, at the initial concept-mapping-training stage, to allow pupils to write headings within the concept maps boxes.
We also found that, at this age, it is challenging for most of the students to identify and extract concepts from a text and then categorize them in hypernyms and hyponyms, and then transcribe the information into a concept map. At the beginning of the concept mapping instruction, we tried to do so. Nevertheless, we decided to allow them to use paragraph headings (generally encompassing adjectives and nouns such as living things, aquatic ecosystems, artificial ecosystems, coral reefs, physical environment). This might also have been one of the reasons why the experimental group identified adjectives as nouns. This group accommodates the information that both categories of words are the same. During the experiment, none of the groups were taught grammar in the Natural Science lessons.
From this study, information that we consider relevant for future research has emerged. It is also perceptible that both groups confuse other lexical and grammatical categories as if they were nouns. On the one hand, the three-month-trained experimental group made the most mistakes in identifying adjectives as nouns. In the previous paragraph, we have exposed one possible cause for this reason, and, in that respect, we think that with some more training, they will be quite more accurate when identifying nouns. On the other hand, the control group did not get confused when discriminating nouns from adjectives. Although, what called our attention most is that this group of control made much more severe mistakes. For example, the vast majority of the participants in this group confused. That is to say, they circled, what are supposed to be easily recognizable lexical and grammatical categories (verbs, adverbs, pronouns, prepositions, conjunctions) with nouns (Figure 7).
Figure 7. Graphic shows a decreasing difference between groups. Three-month-trained pupils improved their marks in 0.04.
To sum up, regarding whether the pupils’ mother tongue (L1), concept mapping has positive effects on text comprehension (Chang, Chen, & Sung, 2002). Concerning the L2, learners having English as a foreign language who use concept maps have achieved much better learning results that those who did not (Mahnam & Nejadansari, 2012). In this regard, language achievement means that the use of concept maps might also help students to develop their receptive (reading & listening) or productive (writing & speaking) skills (0.04). When schoolchildren are starting to develop language skills, during the pre-school/primary education stage, teachers play an essential role in the use and presentation of teaching materials. For this reason, we suggest some future lines of research, especially to those infant and primary teachers whose pupils are immersed in bilingual programs and learn through concept maps. Some longitudinal research regarding the use of concept maps, from infant education on, would be appropriate. Concept maps provide a wide variety of possibilities to do this.
This research was partially supported by Santa María del Prado’s school. We thank primary and high school headmasters at this school for their legal support. We also thank the teaching staff for their assistance during the intervention and data collection
- Ausubel, D. P. (1968). Educational psychology: A cognitive view. New York, Holt, Rinehart and Winston.
- Bacon, T. & Riach, M. (2014). Natural Science: Primary 5. Zaragoza, Edelvives.
- Bentley, K. (2013). The TKT Course CLIL Module. United Kingdom, Cambridge University Press.
- Bussmann, H. (1998). Routledge Dictionary of Language and Linguistics. New York, Routledge.
- Chang, K., Chen, I. & Sung, Y. (2002). The effect of concept mapping to enhance text comprehension and The Journal of Experimental Education , 71 (1), 5-23. Crossref
- De Groot, Annette M. B. (2011). Language and Cognition in Bilinguals and Multilinguals. New York, Psychology Press. Crossref
- Kay, C., Roberts, J., Samuels, M., Wotherspoon, I. (2009). Historical Thesaurus of the Oxford English Dictionary. New York, Oxford University Press.
- Mahnam, L., & Nejadansari, D. (2012). The Effects of Different Pre-Writing Strategies on Iranian EFL Writing International Education Studies, 5 (1), 154-160. Crossref
- Mintzes, Joel J., Wandersee, James H., & Novak, Joseph. D. (1998). Teaching Science for Understanding: A Human Constructivist View. San Diego, Academic Press.
- Novak, J. D. (1988). Conocimiento y Aprendizaje. Los mapas conceptuales como herramientas facilitadoras para escuelas y empresas. Madrid, Alianza Editorial.
- Novak, J. D. (2010). Learning Creating and Using Knowledge: Concept Maps as Facilitative Tools in Schools and Corporations. Routledge, New York. Crossref
- Seuren A. M., Pieter (2013). From Whorf to Montague. United Kingdom, Oxford University Press. Crossref
- Smith F. (2004). Understanding Reading: A Psycholinguistic Analysis of Reading and Learning to Read. New Jersey, Lawrence Erlbaum Associates.