Grammar in primary school EFL textbook activities: a corpus-driven study of their teaching nature and related cognitive implications

One of the fundamental goals of Foreign Language Teaching might be the attainment of implicit knowledge, which underlies oral and written fluency. Textbook activities reflect different explicit and implicit teaching loads following the type of knowledge they are targeted at (explicit, implicit or both). EFL teachers’ implementation of activities can thus help or become a hindrance for the development of implicit knowledge. This study aimed to explore the explicit and implicit teaching nature of 100 activities taken from 10 EFL textbooks used in Spanish primary school teaching. The ultimate goal was to examine whether activities are designed taking into account young learners’ idiosyncratic way of learning. Findings revealed that most primary school EFL textbook activities tend to contain a lower number of activities mostly focused on explicit teaching.


INTRODUCTION
Learning activities represent a fundamental tool for L2 teachers (e.g. Sánchez 2009;Montijano-Cabrera 2014), especially in input-poor contexts of acquisition such as the EFL classroom (Kang 2008). These provide teachers with opportunities to put into practice what has been taught in the class and, with the activities, learners are given the necessary L2 practice for the ultimate attainment of the desired fluency in language use. In an academic environment, most activities are aimed at form-focused instruction (FFI), which covers the explicit and implicit teaching of forms of the L2. Thus, the activity's design is expected to foster mainly the explicit teaching of forms, the implicit teaching of forms or a combination of both. Within primary school EFL teaching, the developmental stage of the learner becomes a key aspect to be taken into consideration. There are specific constraints in child L2 learning that affect the type and amount of attention that younger learners can devote to the conscious processing of rule-knowledge. As a result of this, children may benefit more from implicit teaching actions and materials than from explicit instructional interventions (Baars 1997;Kang 2008; Goorhuis-Brouwer and de Bot 2010; Lázaro-Ibarrola and Azpilicueta-Martínez 2014; Goo et al. 2015;Leow 2015).
Even though there is the assumption that explicit interventions aimed at teaching L2 rule-knowledge in the primary school EFL classroom seem to be less beneficial than implicit pedagogical actions, there is still a strong need to test empirically the correlation between the implementation of certain activities and the derived knowledge gains (Goorhuis-Brouwer and de Bot 2010; Azkarai and Imaz Agirre 2016; Azkarai and Oliver 2016). More specifically, little attention has been paid to (i) child learning in the context of primary school EFL teaching (Lázaro-Ibarrola and Azpilicueta-Martínez 2014), (ii) the method -or approach -to EFL teaching and its impact on learning (Goorhuis-Brouwer and de Bot 2010), (iii) the effects of task repetition on child EFL learning (Azkarai and Oliver 2016) and (iv) the use negotiation of meaning (NoM) strategies among child EFL learners (Azkarai and Imaz Agirre 2016).
The premise that children may benefit more from implicit teaching has led some authors (e.g . Ebbels 2007: 67) to believe that there is "a current move to reduce direct intervention for school-aged children", at least among teachers in charge of younger learners with specific language impairment (SLI). However, it is not only child EFL learners with SLI that might be more frequently exposed to implicit -rather than explicit -teaching actions and materials at school. Normally-developing EFL child learners might also receive more implicit than explicit instruction and materials, as young learners without any language impairment might also be characterized by (Lázaro-Ibarrola and Azpilicueta-Martínez 2014: 3-4): a) being less metalinguistically and sociolinguistically aware b) having developed fewer experiences, cognitive abilities and memory heuristics c) manifesting changes throughout childhood per se, which does not seem to be a homogeneous period, with three clearly differentiated stages (early childhood, 2-7, middle childhood, 7-11, and early adolescence, [12][13][14] In the literature on child EFL learning and teaching, greater emphasis has been placed on the need to conceive learner's preferences regarding the pedagogical strategies and the materials in order to improve the child's learning capacity, his/her memory retention skills, the exposure to stimuli and, overall, the classroom environment (Alizadeh 2011). Similarly, pedagogical decisions such as the use of the L1 in the EFL classroom have been examined to (i) test whether child L2 development is overall improved, (ii) facilitate intake, (iii) enhance comprehension, and (iv) establish an authentic learning environment (Kang 2008: 215). Studies on bilingualism (e.g. Goorhuis-Brouwer and de Bot 2010) have also contributed to the field. Young bilingual EFL learners seem to show better cognitive and metacognitive skills than monolingual children (Goorhuis-Brouwer and de Bot 2010), such as "numeracy, attentional control and creative thinking" (Bialystok 2001(Bialystok , 2007Bialystok, Martin and Viswanathan 2005;cited in Goorhuis-Brouwer and de Bot 2010: 291). Other strands of research within the context of child EFL learning have rather focused on the study of vocabulary acquisition (e.g. Agustín-Llach 2015) and the role of EFL teaching materials (Montijano-Cabrera 2014). Lastly, numerous authors (e.g. Ebbels 2007; Goorhuis-Brouwer and de Bot 2010; Lázaro-Ibarrola and Azpilicueta-Martínez 2014; Azkarai and Imaz Agirre 2015) have expressed their concern about the scarcity of empirical examination on the possible effects of different EFL instructional approaches on child learning, specifically taking into consideration the age and level of proficiency of the subjects being tested. As compared to L2 adult learning, many more changes and transformations might need disentangling in child EFL learning.

Cognitive underpinnings of EFL learning
Within the framework of SLA, one of the ultimate goals of EFL teaching might be the attainment of fluency in the L2 (DeKeyser and Criado 2013;DeKeyser 2015). Explicit and implicit teaching are believed to help the development of the L2. Both have an impact on the type of knowledge attained by the learner. In the literature (e.g. Hulstijn 2002Hulstijn , 2005, it is frequently observed that explicit teaching is closely linked to the attainment of explicit knowledge on the language. Conversely, implicit teaching is frequently associated with the attainment of implicit knowledge. However, for the sake of terminological accuracy and consistency, we would argue that those associations are fairly imprecise. Form-focused instruction covers both explicit and implicit teaching actions, the result of which can be the explicit and/or implicit learning of declarative and/or procedural knowledge on the language (Paradis 2009;Ellis 2009Ellis , 2015Rebuschat 2015). Many different learning outcomes can occur after the implementation of one single instructional action, strategy or material; hence, any imprecise association made for simplification purposes can be misleading or rather erroneous.
What seems true so far is that explicit approaches to EFL instruction are characterized by (i) the use of the L2 as the main object of study, (ii) the frequent delivery of explicit metalinguistic explanations on the language and (iii) the explicit presentation and controlled practice of the target form. Implicit approaches to EFL teaching, on the other hand, are characterized by the use of the L2 as a means of communication, the avoidance of explicit metalinguistic explanations and a reiterative encouragement for free use of the target form. In both scenarios, the teaching action is operationalized through the use of activities 1 (Sánchez 2009). All EFL activities are focused either on explicit teaching, implicit teaching or both on explicit and implicit teaching (Criado, Sánchez and Cantos 2010;Gris 2015). Depending on their focus, the teaching potential that they carry will also be mostly explicit, mostly implicit or mixed. Firstly, activities where forms become the primary focus and meaning is secondary tend to yield a higher explicit teaching load. The following activity could be representative of what might be considered a mostly formfocused activity, since it is primarily aimed at the learning of specific metalinguistic aspects of the FL. Figure 1: Example of an activity mostly focused on the teaching of forms, with a higher explicit teaching potential (Mohamed 2013: 51) Secondly, activities where content is the main goal and the teaching of forms is rather secondary are prone to having a higher implicit teaching potential. Activity 1 in Figure 2 is an example of what could be regarded as a meaning-focused activity. Its main objective is to read a text in order to process it for meaning. Figure 2: Example of an activity mostly focused on the teaching of communicative and/or pragmatic meaning, with a higher implicit teaching potential (Phillips and Shipton 2014: 25) Thirdly, activities where attention to form and content are required tend to carry a mixed teaching potential, that is, partly explicit and implicit. Activity 9 in Figure 3 seems to foster both attention to form -more specifically, questions and answers in the present continuous tense -and to communicative and pragmatic meaning, since the activity encourages learners to put into practice the form they have learnt by asking other classmates. Even though this type of practice might be highly controlled (e.g. repetition drills), engaging in communicative interaction might be conductive to the automatization of knowledge, which, in turn, can foster implicit teaching and learning.

EFL activities for young language learners
Scott and Ytreberg (1990: 1-7) establish a series of characteristics average young learners could be described by. The authors perceive clear differences between two age groups (five to seven-year-olds and eight to ten-year-olds) and provide specific properties that could be attributed to children within each of them. 2 Following Scott and Ytreberg's (1990) distinction, we will put forward some of the fundamental features that activities for child EFL teaching should contain.

Activities designed for five to seven-year-olds should ideally…
• ignore explicit metalinguistic rules and information, since young learners may not be able to understand, interpret or process them consciously; • require the use of the four main skills, namely writing, speaking, listening and reading; • be varied and shift attention between ongoing tasks, as children's attention and concentration span might be limited; • amuse, challenge and reward the learner. Not only do learners like playing games in the classroom, but also they might need to feel that the teacher values them for their interest, work and achievements; • encourage both cooperation and competition; • ask learners about what they are doing, what they have done or heard, as they may like sharing their experiences; • boost the development of important cognitive skills by means of logical reasoning and imagination; • begin to raise learners' awareness of the wide range of intonation patterns that they can use both in their mother tongue and in the L2; • use visual and linguistic elements (e.g. narratives, illustrations); • facilitate teacher-learner interaction to check learner's comprehension. Children may often remain quiet or pretend to understand rather than admit that they do not know something.

Activities designed for eight to ten-year-olds should ideally…
• begin to include basic metalinguistic concepts and brief explanations to offer opportunities for children to infer less salient features in the input; • display extra information on the L2 to boost the interest of more advanced learners; • encourage learners to formulate questions in the L2; • employ a moderate number of visual (e.g. realia, pictures) and aural (e.g. intonation, stress) cues to facilitate the meaning of the language used; • raise awareness of L2 forms by language manipulation through games; • foster cooperative learning as well as competition; • be varied in terms of activity typology, as they might have already developed their own views about which types of activities they prefer or detest to do.

Attention-getting strategies
After listing the features that activities for young EFL learners should ideally contain, there is still a major component to be borne in mind. For learning to actually occur in the classroom -and as a crucial part of the learning construct proposed by Leow (2015) -students have to pay attention, so part of the input can be processed into intake and, later on, stored in the memory as L2 knowledge. Therefore, getting to draw young learners' attention to the target L2 forms and meaning through activities appears to be a crucial endeavour for the EFL teacher (e.g. Alcón-Soler and García-Mayo 2008). A number of researchers (e.g Scott and Ytreberg 1990;Shin 2006;Philp et al. 2008) highlight the importance of not only linguistic input but also visual elements, among other aspects. These can be enhanced through exposure to a variety of stimuli in the classroom. The authors go on to name a series of attention-getting devices in line with the use of adequately designed activities for child EFL learners.

Constant stimuli to keep learners awake
The common belief that young learners' conscious control of their attentional resources is substantially constrained has led to fruitful debates. Children's shorter attention span (as compared to that of adults') could be counteracted by their strong and energetic behaviour, and often by their disposition to discover and learn new things (Philp et al. 2008). As Scott and Ytreberg (1990) and Shin (2006: 5) point out, "words are not enough". Further support of visual components, objects, toys, puppets (or realia, very frequently employed in the Direct Method), body movements and gestures (e.g. as in the Total Physical Response) might be appropriate in combination with activities. The option to include prototypical types of activities from other disciplines or subjects, such as sciences or arts, could also boost learners' attention; for instance, by engaging them in craft activities (e.g. Wright 2008). However, there is the belief that the use of "language-poor activities such as gymnastics, music or handicraft [should be] restricted to a minimum" (Van den Branden 2008: 156), as these may limit the number of potential interactions during which learning can take place.

Variety of materials to avoid monotony
Keeping learners focused throughout the whole lesson might be better achieved with the use of a variety of activities. Moving from one activity to another and not spending -perhaps -more than 10 or 15 minutes with each might improve young learners' attention. In this respect, Scott and Ytreberg (1990: 102) suggest the implementation of consecutive opposite activity types to maintain their attention: quiet versus noisy activities, individual versus pairwork, group work or whole-class activities, teacher-student versus student-student activities, activities promoting different skills at a time (writing, listening, reading, speaking), activities to play with the language versus activities to work with it.

Thought-provoking activities to gain learners' interest
In order to raise attention, interest and awareness of the surrounding world, materials should convey L2 knowledge through different themes. Among the prototypical themes employed to teach young learners are "animals, friends, and family, or units revolving around a story-book" (Shin 2006: 4). As children grow older, other topics involving a slightly higher degree of complexity appear in the curriculum, such as the environment, citizenship or shopping. Teaching in themes might facilitate a wider and more familiar context for learners to pick up the language (Philp et al. 2008). Of course, the themes should be suitable to the level of proficiency of the learners and their use coherent within the thread of lesson and with the learning objectives in the curriculum. Similar content-areas might offer opportunities for the EFL teacher and the learners to elicit or recycle prior knowledge (e.g. vocabulary seen in previous units).

Habit-formation to build a comfortable learning environment
Predictability can build a positive and comfortable learning environment by lowering learners' anxiety and achieving better concentration. For the EFL teacher, following specific "routines in the classroom can help to manage young learners" (Shin 2006: 5). Classroom routines happen every day and these give the learners opportunities to develop their L2 (Cameron 2001) as long as the EFL teacher implements -or designs -activities that require interaction. Attention-getting devices such as thematic class divisions or class-section labels said aloud by the teacher (e.g. It's speaking Friday!) can be useful to form habits and routines every week (Cekaite 2008).

L1 use to support L2 comprehension
The attempts to reduce EFL teacher interventions in the L1 may not always be favourable for all students in the classroom, let alone for those with lower levels of proficiency (e.g. White 2008). The L1 can become a useful teaching instrument, especially when trying to convey complex words or expressions. Also, it can foster motivation among those who find communication-in-English-alone difficult to understand or follow. Resorting to the L1 in such circumstances can "save time for the target language that is actually within students' reach rather than spending time trying to make a difficult word or expression comprehensible in English" (Shin 2006: 6). This however should not entail that the L2 is replaced by the L1 at all times. After conveying clarifications in the L1, the EFL teacher may wish to recast the same in the L2.

OBJECTIVE
The objective of this exploratory study was to uncover the teaching nature (explicit, implicit or mixed) of 100 activities taken from 10 popular ELT textbooks used in primary school teaching in Spain -to my knowledge, no previous related studies have been carried out. The ultimate aim was to examine whether activities per se are designed taking into account the possible constraints in child L2 learning.

METHOD
In order to ascertain the teaching nature of 100 activities, a series of subtasks were necessarily undertaken. The first subtask implied the compilation of a sample of primary school EFL activities from a variety of popular primary school EFL textbooks (see Sections 3.1 and 3.1.1). The second subtask involved the random selection of 100 activities extracted from those same textbooks (see Section 3.1.2). The third subtask entailed the application of the measuring instrument in order to discern the explicit and/or implicit teaching potential of the selected activities (see Section 3.2). Last, the fourth subtask was aimed at determining the degree of inter-judge reliability to validate the process of activity categorization (see Section 3.2.2). A more thorough account of each of the procedures undertaken is displayed below.

Sample of textbook activities
A complete body of 36 popular primary school EFL textbooks in Spain (see Appendix 1) was employed for the compilation of a sample of real teaching activities. Both the student book and the workbook were susceptible to being selected. At no point during the investigation process was copyright infringement intended. The analysis of textbooks and activities was only carried out for strict research purposes. All the textbooks and workbooks were published by some of the most popular ELT textbook editorial houses in Spain, namely Oxford University Press, Cambridge University Press, Macmillan, Pearson Longman, Pearson and Richmond Publishing.

Selection of textbooks
From the complete body of 36 popular Spanish primary school EFL teaching textbooks, we made a random selection of a sample of 10 elementary (A1, A2) textbooks dating from 2004 until 2015 used in the 2nd and 3rd cycles (years 3rd to 6th) of primary school. For randomization purposes, one of the criteria followed for the selection of textbooks was that different editorial houses should be present in the sample. After the selection process, there were 5 ELT textbooks by Oxford University Press, 2 textbooks by Macmillan, 2 other textbooks by Longman and, lastly, 1 textbook by Richmond. All the activities (including the extra activities in the revision units across the textbooks) in each of the 10 textbooks below were compiled in the sample. By the end of the compilation process, the complete sample contained a total number of 1,952 activities. For the selection of activities within the sample, two major steps were required. The first step entailed the random selection of one complete unit from each of the 10 textbooks in Table 1. The second step was the extraction of 10 activities from each of those individual units. Activity extraction, however, was not fully random. This decision was taken to maintain activity sequencing. Having opted for a random selection of activities might have become a threat to the validity of the process of activity categorization; thus, it was more convenient to select 10 activities in a row from each of the units in the textbooks. By following these two criteria, we randomly selected 100 authentic primary school EFL textbook activities from the nearly 2,000 activities compiled in the 10 randomly chosen textbooks. 10 activities were serially extracted from each of the 10 textbooks mentioned above in Table 1.

Procedure for the analysis of the activities
For the compilation of the sample of activities, an Excel sheet was used. A series of metadata were included in order for us to be able to locate the source of each activity at all times. From left to right in Figure 4 below, the dataset included information about: (i) the name of the textbook, (ii) the category of edition (SB for student book and WB for workbook), (iii) the unit and the page where the activity would be found, (iv) the activity number, (v) the type of activity regarding the skills promoted, namely writing, listening, reading and speaking, and specific components within different areas of EFL teaching, such as grammar, vocabulary, spelling or pronunciation (it should be noted that different pedagogical strategies could be employed in some activities); and (vi) the activity's heading.

Categorization and coding of the activities
Two co-authors individually examined all the 100 activities. Firstly, they read each individual activity's instruction in the sample. Secondly, they examined the activity and its sequencing in the EFL textbook. Thirdly, the activities were categorized regarding their teaching potential. As can be observed in Figure 4 (right column), the code <0> was assigned to those activities where there was a full attention to form and where meaning was secondary. The code <100> was employed in the case of activities with a full focus on the meaning and where form was a secondary aim, if at all. Lastly, the code <50> was assigned to activities that required the learner to pay attention both to form and meaning.

Inter-rater reliability
The convenient statistical procedure to test the inter-judge reliability in activity categorization was Cohen's Kappa, as no more than two judges were involved in this analysis (Fleiss' Kappa is used when three or more judges take part in the analysis).
In Table 2 below, data derived from raters' individual activity categorization are crisscrossed to determine the coincidences and mismatches. As can be seen, there was a full coincidence between raters with regard to their analysis of mostly explicit (<0>) activities. There were 33 out of 37 possible matches in the categorization of activities whose teaching load was mostly implicit (<100>) and there were 40 out of 43 possible matches between raters in the categorization of activities with a mixed potential (<50>).
Overall there were 93 out of 100 matches between raters 1 and 2, which suggests that inter-judge reliability was very high. After the application of Cohen's Kappa, the exact inter-judge reliability value was revealed (x=0.891), which proved the high degree of validity of the measuring instrument and the raters' analysis.

RESULTS
In this section, results will be analyzed from three different perspectives. First, by examining the typology and amount of activities overall present in the sample collected, I will provide a general overview of the results (Section 4.1). Second, results from each individual textbook will be shown in a chronological order (regarding their date of publication) to determine their type and amount of activities (Section 4.2). Third, results will be merged into a cluster tree to examine specific trends in the data more closely (Section 4.3).

General results
Based on the categorization of the 100 activities regarding their explicit and/or implicit teaching nature, it can be observed (see Figure 5) that there is a higher presence of activities with a mixed teaching potential; that is, those labelled as <50> (N=43). The presence of activities whose teaching potential was mostly implicit -or <100> (N=37) -was also noteworthy. Not surprising, however, was the fact that fewer activities with a higher explicit teaching potential -or <0> -were found in the sample (N=20).
The tendency to include a lower number of activities with a higher teaching potential in textbooks seems to be in line with the proposals and guidelines found in the literature on child EFL learning (Scott and Ytreberg 1990;Goorhuis-Brouwer and de Bot 2010;Lázaro-Ibarrola and Azpilicueta-Martínez 2014). At this point the reader should not misinterpret the under-representation of activities with a higher explicit teaching potential in textbooks as an indicator of biased teaching and/or learning. Indeed, explicit teaching occurs in most activities -not alone, however -but in combination with implicit teaching, in activities with a mixed potential (N=43). Indeed, reducing the amount of activities with a higher explicit teaching potential to a minimum in child EFL textbooks may overall suggest that: • the implicit and the explicit inductive modes of teaching are preferred over the explicit deductive mode in the teaching of the L2 forms; • teaching is primarily focused on the implementation of activities that foster the implicit -or explicit inductive -practice of skills (writing, speaking, listening and reading) across differentiated periods of childhood (Scott and Ytreberg 1990;Goorhuis-Brouwer and de Bot 2010); • the reinforcement of specific components within different areas of EFL teaching (vocabulary, grammar, spelling and pronunciation) is most times present in activities, but subtly integrated in a communicatively-oriented task; 3 • the presence of at least some key metalinguistic information is necessary to start to develop basic notions; • several teaching strategies or modes are combined in each activity in order to amuse younger learners and better focus their attention on the learning objectives; • encourage both individual and work group rather than individual work alone (the latter might often be promoted in explicit deductive activities) (Scott and Ytreberg 1990: 2-3).

Individual results for each textbook
The results for each individual textbook are exposed in this section. Figure 6 below displays the textbooks chronologically ordered and the type and amount of activities found within each of them.
First, English Adventure 3 (student book), Find Out 3 (student book) and Big Surprise 5 (student book) contain a larger number of activities with a higher implicit teaching potential. English Adventure 3 and Find Out 3 are EFL textbooks specifically used in the 3rd year of primary school (2nd cycle), when students' age is assumed to range between 8 and 9. Big Surprise 5, however, is meant to be used in the 5th year of primary school (3rd cycle), for students of ages 10 to 11. The distribution of activities with regard to their typology and potential in these three textbooks suggests that very little to no space at all is given to activities with a higher explicit teaching load. Even though it is recommended that explicit teaching activities -to explicitly focus on the L2 forms -begin to appear at the age of 8 (Scott and Ytreberg 1990;Goorhuis-Brouwer and de Bot 2010), some textbooks (e.g. English Adventure 3, both the student book and the workbook) do not seem to include them, at least in all units.
Second, the textbooks found to contain a greater number of activities with a mixed teaching potential were Find Out 3 (workbook), Mega Zoom 3 (student book), Oxford Rooftops 3 (workbook) and Great Explorers 6 (workbook). The first three are used in the 3rd year of primary school (2nd cycle), when students' ages range between 8 and 9. The last one (Great Explorers 6) is designed for students in their 6th year of primary school (3rd cycle), with ages 10 to 11. Here, I invite the reader to reflect on the fact that three out of these four books are workbooks. Based on mere experience, workbooks can give the impression that explicit and controlled language practice is the preferred mode. However, after empirical examination and in the light of the evidence obtained, this belief should not be taken for granted. Indeed, only one of the textbooks containing a higher number of activities with a mixed teaching potential is a student book: Mega Zoom 3. All this may indicate that not only the activities found in the student books, but also the ones designed for the workbooks abide by and follow the principles of child EFL teaching and learning. In other words, not only the student book but the workbook in the 2nd and 3rd cycles of primary school contain fewer activities with a higher explicit teaching potential. Instead, EFL material designers seem to take into account children's idiosyncratic way of learning and they may attempt to incorporate more communicatively-oriented activities where forms are taught implicitly or in an explicit inductive way.
Noteworthy are the cases of English Adventure 3 (workbook), Great Explorers 6 (student book) and Big Surprise 6 (workbook). Firstly, in English Adventure 3 (workbook) and Great Explorers 6 (student book), the number of activities within the categories mostly implicit (<100>) and mixed potential (<50>) is the same. Secondly, the case of Big Surprise 6 (workbook) is worth highlighting, as it may stand as an exception to 'the rule'. The teaching potential of 80% of the activities found in Big Surprise 6 (workbook) was mostly explicit (<0>), showing no instances of mostly implicit (<100>) activities. The findings obtained from the analysis of the typology and amount of activities within Big Surprise 6 (workbook) reveal that this book does not seem to follow the principles and guidelines for child EFL teaching and learning established in the literature (Scott and Ytreberg 1990;Goorhuis-Brouwer and de Bot 2010;Lázaro-Ibarrola and Azpilicueta-Martínez 2014). A very similar scenario occurs in Find Out 3 (workbook), where 80% of the activities are aimed at both explicit and implicit teaching (mixed potential, <50>) and, again, there are no activities within the category mostly implicit (<100>).
In contrast to the aforementioned cases where there was a complete absence of mostly implicit (<100>) activities, English Adventure 3 (both the student book and the workbook) do not include any activities with a mostly explicit (<0>) teaching potential.

Similarities and differences between textbooks
In order to detect and statistically verify the interrelation between the data regarding the type and amount of activities in the different textbooks, a cluster analysis was carried out. This type of statistical inspection is useful to ascertain the possible trends in the data in a visual way. The output of a cluster analysis is a dendogram -or cluster -tree (see Table 4), where the different branches depict numerical tendencies that would probably be difficult to notice if mentally computed. Table 4 below shows the type and amount of activities regarding their teaching nature in the different textbooks from the sample. The cluster tree represents the interrelationships between them with 'tree branches'. Cases 4, 9, 10, 5 and 8 in the biggest branch at the bottom of the cluster tree -corresponding to Find Out 3 (workbook), Great Explorers 6 (student book), Great Explorers 6 (workbook), Mega Zoom 3 (student book) and Oxford Rooftops 3 (workbook) -mostly contain a higher number of activities with a mixed potential (those coded as <50>). Cases 5 and 8 -shown at the very bottom of the dendogramhave the same number of activities within each of the categories mostly explicit <0>, mixed potential <50> and mostly implicit <100>.
Cases 1, 2, 3 and 6 shown in the cluster tree -corresponding to English Adventure 3 (student book), English Adventure 3 (workbook), Find Out 3 (student book) and Big Surprise 5 (student book) -mostly contain a higher number of activities primarily focused on implicit teaching (coded as <100>). Additionally, with regard to the textbooks containing a higher quantity of activities whose teaching nature was mostly implicit, cases 1 and 2 (the first branch of the dendogram) point to two instances where there were no activities mostly focused on explicit teaching (coded as <0>).
Lastly, the isolated branch in the cluster analysis -case 7, corresponding to Big Surprise 6 (workbook) -stands for the only instance from the sample analyzed where a textbook contained a larger number of mostly explicit (<0>) activities. The fact that no activities within the category mostly implicit (<100>) were observed should also be noted.
As a result of this data-driven inspection, it could be critically argued that the most suitable, integrative and least biased textbooks for the courses of primary school where these are implemented are Find Out 3 (student book), Great Explorers 6 (student book), Great Explorers 6 (workbook), Mega Zoom 3 (student book) and Oxford Rooftops 3 (workbook).

DISCUSSION
The objective of the present exploratory study was to uncover the teaching nature of a sample of primary school EFL teaching activities. Recent research in the field of child EFL learning has examined and evaluated: (i) learners' interaction during task performance (cf. Lázaro-Ibarrola and Azpilicueta-Martínez 2014; Azkarai and Oliver 2016), (ii) approaches to EFL teaching under different instructional settings and learner conditions (cf. Ebbels 2007;Kang 2008), (iii) the role of the textbook in the EFL classroom (Montijano-Cabrera 2014), (iv) vocabulary development across learners and throughout different stages in primary school (Agustín-Llach 2015) and (v) the teaching action and its effects on learners' proficiency (Goorhuis-Brouwer and de Bot 2010). The findings in this study revealed that there is the tendency in primary school EFL textbooks to give prominence to activities partially or mostly focused on communicative and/or pragmatic meaning, whereas metalinguistic awareness is not very frequently pursued. This tendency could be strongly supported by evidence in the literature, taking into account the specific characteristics of child EFL teaching and learning, thoroughly portrayed at the beginning of this paper. However, most studies conducted seem to show common weaknesses regarding their methodology (e.g. number of participants, task, instrument for data collection, among others). These may constitute a major threat to their internal and external validity. But for one study (Goorhuis-Brouwer and de Bot 2010), the majority of the papers reviewed tend to demonstrate a lack of solid samples of participants and a scarcity of validated and reliable measures for data collection and analysis (e.g. Ebbels 2007; Kang 2008; Lázaro-Ibarrola and Azpilicueta-Martínez 2014; Montijano-Cabrera 2014; Azkarai and Oliver 2016). For instance, whereas more than 300 subjects completed a highly reliable standardized test in Goorhuis-Brouwer and de Bot's (2010) study, the other works seem to display a rather poor construct validity with data obtained from smaller samples of informants (cf. Ebbels 2007;Kang 2008, with three and one participants, respectively).
Task selection processes might have constituted another threat to the validity of some studies. For example, in two similar studies aimed at examining learners' interaction, a guessing game was implemented to observe which negotiation of meaning strategies were chosen by the learners (e.g. Lázaro-Ibarrola and Azpilicueta-Martínez 2014; Azkarai and Oliver 2016). In the latter, the fact that activity selection was not counterbalanced was even acknowledged by the authors. Lastly, other studies like Montijano-Cabrera (2014) are not empirically-driven.
As far as the results are concerned, most research on primary school EFL teaching has reported on the lack of empirical evidence to support CLIL for the attainment of certain skills, such as the development of vocabulary (Agustín-Llach 2015) or effective strategies to interact (Azkarai and Oliver 2016), at earlier stages in school. This may suggest that younger learners who are highly exposed to implicit teaching interventions and activities (e.g. CLIL subjects) in the context of primary school EFL teaching in Spain might not always attain better results than those who attend ordinary EFL classes alone. In other countries in Europe, within other instructional settings and with different language curricula and policies, the outcome of CLIL programs might nonetheless be more effective, as reported by Goorhuis-Brouwer and de Bot (2010). Indeed, Agustín-Llach (2015) points to the fact that the young age and low proficiency of learners in the studies conducted in the Spanish context might have constituted a potential block for a distinct outcome.
What seems true so far is that age, instructional settings and activities are three key factors in the context of primary school EFL teaching and learning. Since most research conducted -to my knowledge -has not dealt with the dimension of the typology and/or teaching nature of the learning activities employed in the classroom, the present study was aimed at fulfilling that research niche. This study should also be potentially motivating for other researchers to seek validated and more reliable measures for activity selection if future research on the field of primary school EFL teaching is carried out, or replications to the studies reviewed are conducted.

CONCLUSIONS AND PEDAGOGICAL IMPLICATIONS
Activities seem to play a fundamental role in the context of primary school EFL teaching and learning. Inasmuch as teachers are responsible for their use in the classroom, activity selection and implementation should become a key aspect in the research arena of child EFL learning. Since there are potential pedagogical and cognitive implications derived from the use of activities within instructional settings, EFL instructors should not ignore the nuances across different typologies of learning activities, let alone at earlier stages of education and, more specifically, with younger learners. Being aware of the teaching nature of different activities might be a starting point to improve teaching interventions and attain more successful teaching and learning experiences. Following this rationale, the objective of the present exploratory study was to examine the teaching potential (explicit, implicit or mixed) of a sample of 100 activities taken from 10 popular primary school EFL textbooks in Spain.
Most activities found in primary school EFL textbooks tended to display a higher implicit teaching potential or a mixed one, that is, partly implicit and explicit. The tendency in textbooks to contain a lower amount of activities mostly focused on the teaching of forms -and where the focus on meaning was secondary -seemed to match younger learners' cognitive needs. However, this seemingly positive finding suggests that EFL material designers might not be totally aware of the techniques and strategies that best suit younger learners. Indeed, the presence of a considerable amount of explicit teaching activities in textbooks might suffice to support the claim that textbooks should contain a more adequate proportion of activities aimed at teaching EFL to younger learners. Notwithstanding the disproportionate presence of explicit teaching activities in some textbooks, it seems true that research is going in the right direction and studies like the present one might stand good chances of being beneficial to that end. Put it another way, this piece of research should be taken into account by EFL material designers to improve the adequacy of primary school EFL textbooks for younger learners.
Some of the future lines of research might entail: (i) the compilation of a bigger sample of activities and the analysis of the teaching potential (explicit, implicit or mixed) of a higher amount of activities; (ii) the study of tendencies in the typology of activities within the categories mostly explicit, mostly implicit and mixed potential; and (iii) the analysis of motivation as related to the different typologies of activities and their corresponding teaching potentials.
With regard to the pedagogical implications of this study, it could be argued that there is a strong need to provide specialized training on the detection and analysis of the teaching potential of activities to current primary EFL teachers and teacher trainees. It is essential that, at all times, the practitioners are aware of the different teaching potentials of the activities they use in class (e.g . Freeman 1989;Kelly 2017), so that their pedagogical actions do not clash with the children's idiosyncratic way of learning. In other words, teaching should always facilitate and optimize learning.
It could also be argued that EFL teachers should acquire a more critical attitude towards the materials that they handle in the classroom. Instructors should be highly aware of the possible learning outcomes of different teaching interventions and the use of a wide array activities, techniques and strategies. Therefore, developing a critical mindset in this respect should serve as model to choose, filter or even discard teaching materials taking into account variables such as age, learners' profile or type of instructional settings at any time.
Last, one very positive pedagogical implication derived from the application of a critical attitude towards EFL teaching material use is the ultimate improvement in terms of quality and amount of input that the learner will eventually be exposed to.
ACKNOWLEDGMENT This work was supported by the Spanish Ministry of Economy, Industry and Competitiveness [Research project reference FFI2013-44979-R]