Moroccan EFL Public High School Teachers’ Perceptions and Self- Reported Practices of Assessment

This study aims to investigate the perceptions and self-reported practices of Moroccan EFL public high school teachers towards traditional and alternative assessment. The data were collected from 51 teachers in Northern Morocco using a self-developed online questionnaire. The questionnaire items about teachers’ perceptions and self-reported practices were valid and both their data and sampling were acceptable for factor analysis of three subscales (traditional assessment, alternative assessment related with assessment as learning, and assessment for learning), and all scales proved to be reliable. Based on the three research questions, the study yielded the following results: (1) Teachers perceived the objectives of alternative assessment to be significantly more important than those of traditional assessment. (2) Based on their self-reported practices, teachers mainly used traditional assessment methods more often than alternative assessment methods associated with assessment as and for learning. (3) When comparing teachers’ perceptions with their self-reported practices, we found that teachers’ perceptions regarding traditional assessment matched their practices; while the majority of teachers admitted that they found alternative assessment important even though they did not often use it in order to support students to be able to reflect on their own learning or to enhance their performance in the learning process. Thus, these findings are significant for researchers, teachers, and educators to help them reconsider their perceptions of alternative assessment and how they should be enacted in practice with the aim of resolving the mismatches found in this study.


Introduction
Assessment has been one of the biggest areas of interest for scholars and practitioners in a multitude of fields of language teaching and learning. Traditional assessment (TA) has been applied in schools and for job recruitment and qualifications for many years. Recently, however, other forms of alternative assessment (AA) have been developed to meet both teachers' instructional objectives and students' needs in terms of their longterm progress evaluation. Despite the globally increasing interest in AA practices (Stognieva, 2015), the Moroccan educational system still grapples with the use of TA methods, which have often resulted in pass-fail decisions (Bouziane, 2017;Ouakrime, 2000). The high-stakes English examinations at the Baccalaureate level in Morocco have been the major criterion used to measure students' language skills despite the opportunity for English teachers to implement the recommended forms of formative assessment (e.g., portfolio, project, selfand peer-assessment) as prescribed in the English language guidelines (Ministry of National Education, 2007). Owing to such a strict reliance on final examinations, it cannot be claimed that teachers are free to switch from TA to AA, and therefore this has a direct impact on students' motivation for learning and on teachers' preparation of instructional materials, known in language testing and assessment research as the washback effect (Ghaicha & Oufela, 2021;Green, 2013). Thus, Moroccan EFL teachers are required, on the one hand, to administer and prepare their students for this exam, and, on the other hand, they are expected and encouraged to use AA methods, which is, in practice, negatively influenced by their time constraints, class sizes, and lack of training (Ghaicha & Omarkaly, 2018). The need for the more purposeful and frequent use of AA to increase the efficiency of English learning and teaching is warranted, especially on the part of the school administrations, which can encourage and support teachers. To address the problem, this study investigated the assessment ABDERRAHIM MAMAD, TIBOR VIGH situation in Moroccan public high schools in which EFL is introduced to students during the second year (after one year of English study in middle school). To achieve the purpose of this investigation, examining teachers' perceptions and self-reported practices of both TA and AA is warranted to help identify the extent to which their perceptions are enacted via specific classroom-based practices regarding the use of different assessment methods.

The Framework of Assessment for/as/of Learning
Within the framework of assessment, there is considerable inconsistency surrounding the meanings of classroom-based assessment (CBA). Rea-Dickins (2007) considered it to be effective formative language assessment where "learner performance is analysed in terms of learning goals and instructional processes rather than a finished product" (p. 507). In terms of practices and procedures, Davison and Leung (2009) viewed CBA as linked to and conceptualised as teacher-based assessment (TBA). The latter shares many characteristics with assessment for learning, which is often identified as a part of formative assessment and defined usually in opposition to traditional examinations, which are primarily designed for the purposes of selection and accountability (Davison & Leung, 2009;Lan & Fan, 2019). In contrast, Hill and McNamara (2012) defined CBA from a holistic perspective involving both formative (assessment for/as learning) and summative assessment (assessment of learning).
To provide definitions for these types of assessment, three main concepts were developed originally by Black and William (1998). The first concept is assessment for learning (AfL), which is defined by Gan, Liu, andYang (2017, p. 1126) as "a process in which teachers seek to identify and diagnose student learning problems, and provide quality feedback for students on how to improve their work" so that they can meet the intended outcomes (Davidson & McKenzie, 2009). The second one is assessment as learning (AaL). It is based on learners' reflections on evidence of learning through their ability to build knowledge of themselves as autonomous and independent learners (Briggs, Woodfield, Martin, & Swatton, 2008;McDowell, Wakelin, Montgomery, & King, 2011). AaL also requires teachers' involvement in supporting students' learning by providing them with feedback regarding their performance (Earl & Katz, 2006). The third concept, assessment of learning (AoL), is used to "determine and report on student achievement at the end of a learning cycle" (Lan & Fan, 2019, p. 112).
Essentially, there are two basic modes of CBA that might be interrelated with other types of assessment: traditional and alternative. The former, linked to summative assessment (AoL), is designed to determine the extent to which learners meet the instructional goals and outcomes, as well as to confirm what and how much they know at the end of the learning process (Davidson & McKenzie, 2009;Earl & Katz, 2006;Nasab, 2015). On the other hand, the latter, associated with formative assessment, is meant to find information not only about what students know and can do with what they have learned (AfL), but also how they can reflect on and guide their own learning progress (AaL) (Al-Mahrooqi & Denman, 2018;Brown & Hudson, 1998).

Traditional and Alternative Assessment
Whereas AA requires students to be able to demonstrate what they have learned in the classroom through meaningful tasks that replicate real-life situations (Phongsirikul, 2018), traditional test practices direct teachers to assess students' language skills through summative tests and final examinations. They are also designed to provide information on students' achievements rather than on their learning strategies, interests, and motivations (Nasab, 2015). Because traditional methods do not assess reflective/critical thinking, self-, and peer-evaluation (Stognieva, 2015), AA has become a prominent approach that combines learning, instruction, and evaluation (Al-Mahrooqi & Denman, 2018;Anderson, 1998). According to Bolshakova (2015) and Brown (2004), some of the main characteristics of AA revolve around developing students' cognitive ability to create, produce, or perform performance-based assessment tasks that reflect real-life situations. Thus, the aim of AA is to support the development of learners who are able to link previous information with new experiences in the real world. In contrast, in a context where only TA is used, this assessment, as argued by Shams and Tavakoli (2014), is considered undesirable because of its focus on rote learning, on the reproduction of information on exams, and on judging students' performance based on scores (grades). However, this argument does not neglect the use of TA, but effective CBA should also include alternatives. The latter might contain tests and other non-standardised methods, which, like TA, need to be constructed in responsible, rigorous, and principled ways (Brown, 2004) by teachers to assign scores for or make judgements on students' final products (Brown & Hudson, 1998).
As another comparison, Aksu Ataç (2012, p. 11-12) distinguished between standardised testing, which is a form of TA, and authentic assessment, which is referred to as AA (Hamayan, 1995;Monib, Karimi, & Nijat, 2020). The former regards testing and instruction as separate activities, focuses on lower-order knowledge and skills (recall or recognition of facts, ideas, and propositions), forbids students to interact, and provides normreferenced feedback through comparisons with others' results (Chan & Liam, 2010;Omari, Moubtassime, & Ridouani, 2020). On the other hand, the latter considers assessment to be an integral part of instruction (Al-Mahrooqi & Denman, 2018), focuses on higher-order learning outcomes and higher-level thinking skills such as analysis, synthesis, reflection, creative thinking, problem-solving, and the application of information and knowledge (Al-Mahrooqi & Denman, 2018;Brown & Hudson, 1998), and, in accordance with principles of criterion-referenced testing, compares learners' performances with aims and proficiency levels. AA also encourages collaborative and autonomous learning (Nasab, 2015), and compares learners to their previous achievements when providing self-referenced feedback about their improvement (Chan & Liam, 2010;Omari et al., 2020).
Despite the differences between TA and AA, it is productive and advantageous if teachers do not have to choose between them. Likely, some mix of the two will best meet the teachers' or students' needs to have deeper insights into the learners' main strengths and weaknesses (Babni, 2019). For this latter aim to happen, Nasab (2015) argued that "teachers must be attentive to the diverse ways of assessment and not to rely too heavily on a single method of assessment" (p. 175). Thus, effective CBA involves the dynamic use of multiple sources of information collected through TA and AA tasks (Baranovskaya & Shaforostova, 2017).

The Role of EFL Teachers' Perceptions and Self-Reported Practices in Assessment
The term 'perceptions' is defined as uniquely individualised experiences, mental and personal constructions, assumptions, and propositions (McDonald, 2012;Richards & Schmitt, 2010) which, in the EFL context, influence teachers' judgements and decisions (Kirkgoz, Babanoglu, & Ağçam, 2017). Whereas self-reported practices are indicators for "which teaching practices were used during a lesson and how often they were employed […] by teachers […]. Self-reporting is intended to assist teachers to assess their own classroom practices" (Richards & Schmidt, 2010, p. 476). In the context of assessment, investigating teachers' perceptions is important to the understanding and implementation of CBA (Sach, 2012). For example, if "the teachers accept or have positive perceptions on alternative assessment, they will surely support the assessment and make sure the AA succeeded in reality" (Nasri, Roslan, Sekuan, Bakar, & Puteh, 2010, p. 40). Self-reported practices also have meaningful alignment with perceptions, beliefs or/and conceptions (Brown, 2009;Brown, Chaudhry, & Dhamija, 2015). Thus, the extent to which teachers' perceptions match their practices has a great influence on their mode of assessing their students' performance (Hakim, 2015). However, when using CBA effectively, teachers' assessment literacy plays an important role because it connects assessment quality with student achievement (Ashraf & Zolfaghari, 2018). As a wider definition, this construct also includes theoretical and practical knowledge, skills, and abilities, which are "required to design, develop, maintain or evaluate, large-scale standardised and/or classroom-based tests, familiarity with test processes, and awareness of ABDERRAHIM MAMAD, TIBOR VIGH principles and concepts that guide and underpin practice, including ethics and codes of practice" (Fulcher, 2012, p. 125).
Related to TA and AA, some previous studies conducted in different contexts, focused on several assessment components or characteristics when analysing teachers' perceptions and their self-reported practices. The findings of Phongsirikul's (2018) study demonstrated that both teachers and students accepted TA more than AA. Kirkgoz et al. (2017) found that pen-and-paper exams, performance tasks, in-class observation, quizzes, and project assignments are the most frequently used assessment types, whereas oral exams and presentations are the least employed types in assessing learners' performance. Nasri et al. (2010) found that teachers have positive perceptions of AA, especially as a way of promoting students' self-confidence and involvement in learning as well as for developing their critical and creative thinking skills. In the Moroccan context, most of the recent studies (Babni, 2019;Benzehaf, 2017;Ghaica & Omarkaly, 2018;Ghaicha & Oufela, 2021) targeting the EFL high school context have mainly explored Moroccan teachers' perceptions and practices related to AA. Since it is not typical in Morocco to find a common investigation of teachers' perceptions and self-reported practices towards TA, AA, and their related assessment concepts (AoL/AfL/AaL), a consideration of such duality is warranted. Therefore, the purpose of this study is to describe teachers' perceptions and self-reported practices with both TA and AA in Moroccan public high schools. To achieve this objective, the following questions were addressed: 1. What are Moroccan EFL high school teachers' perceptions about TA and AA objectives? 2. What are Moroccan EFL high school teachers' self-reported practices regarding the application of both TA and AA methods? 3. What is the relationship between Moroccan EFL high school teachers' perceptions and their selfreported practices?

Methodology Participants
The present study was conducted in Morocco with a focus on Moroccan EFL public high school teachers. An exploratory quantitative study using a survey method was designed, first, to be able to analyse the similarities and differences between teachers' perceptions about TA and AA associated with AaL and AfL in terms of their objectives and methods, and second, to compare the perceived importance of the specific type of assessment objectives with its counterpart regarding the application frequency of its methods. Thus, data were randomly collected through an online questionnaire. The study involved 51 EFL teachers at different public high schools in the northern part of Morocco, mainly in Rabat, Casablanca, Tangier, Meknes, and Fes. Table 1 contains the main characteristics of the sample based on the background information collected. The number of males exceeds the number of females. Teachers between the ages of 26-30 and 31-35 are the dominant sub-sample. Respondents with 1-5 years of teaching experience are the most significant participants. As for the number of students in a class, most teachers teach around 31-35 students or even more. Therefore, mostly young, novice teachers, who have been teaching large groups of students, participated in our study.

Instrument and Procedure
In the current study, a self-developed, three-part questionnaire was used (see Appendix). The first part asks general background information, allowing for an appropriate description of the sample; thus, the variables (Table 1) will not be considered in data analysis and interpretation.
To meet the first purpose of the study regarding teachers' perceptions about TA and AA objectives, the next part of the questionnaire was designed based on the aims and main characteristics of these assessment modes as well as the concepts of assessment for/as/of learning. When developing the questionnaire items, 13 total statements were formulated. Out of these, four items are related to TA and which also belong to AoL (e.g., Assessment should assign grades to students.). The other items describe AA in two ways: five statements cover AaL in terms of how teacher assessment can support students' assessment, performance, and learning (e.g., Assessment should provide students with the information they need to guide and improve their own learning.). The remaining four items represent AfL (e.g., Assessment should focus on what students can do with what they have learned.). For each item, the teachers had to decide on the extent to which they agree/disagree with the statements formulated. A five-point Likert scale was used ranging from one (strongly disagree) to five (strongly agree). A scale with an odd number was chosen because it allowed teachers to express their neutral positions.
To fulfil the second purpose of the study, the aim of preparing the last part was to explore teachers' selfreported practices behind the application of TA and AA methods. Therefore, this part required teachers to rate how frequently they use 18 assessment methods. Nine of them were related to TA (e.g., multiple-choice tests, essays, short-answer tests, etc.), six items included AA methods related to AfL (e.g., reports, conferences, projects, etc.), and three items focused on AA methods linked to AaL (e.g., self-and peer-assessment). In the case of all methods, a five-point Likert-type scale was offered ranging from one (never) to five (always). Since these points are approximately the same distance from each other, they can be interpreted on interval scales.
Before using the designed questionnaire, the clarity of its items was tested via a pilot study by collecting remarks from a group of teachers, which were taken into consideration while finalising the questionnaire. The latter was distributed online and anonymously completed by the involved respondents. As the third aim of our study was to explore the relationships between teachers' perceptions and their self-reported practices, first the validity and reliability of the questionnaire items related to these two measured dimensions were examined to ensure that the items are relevant to the research focus, and before analysing them at scale level, the extension of the results are reliable and valid. Thus, the purposes of using exploratory factor analyses were to identify and compare the empirical structure of the variable system with the theoretical structure and to reduce the data set to a manageable size while maintaining the original information of the items as much as possible (Field, 2009;Pituch & Stevens, 2016). Therefore, principal component analysis (PCA) with Varimax rotation was performed to determine the contribution of each variable to the factor structure and to create composite scores. The reliability of the scales was also examined by calculating the values of Cronbach's alphas.
As for the factorability of the 13 questionnaire items on teacher perceptions, the Kaiser-Meyer-Olkin measure of sampling adequacy (KMO) was .54, which, as recommended by Kaiser (1974), was mediocre and above the minimum acceptable value of .5. The KMO values for the individual items ranged between .38 and .76. Furthermore, Bartlett's test of sphericity was significant, χ 2 (78) = 177.24, p < .001, and confirmed that the correlation coefficients between the items were sufficient for PCA. Additionally, each item correlated significantly, .23 ≤ r ≤ .63, p < .05, with at least two other variables. Likewise, communalities were above the minimal acceptable limit of .3 in all cases (ranging from .30 to .74, the average was .53). The number of factors was determined based on the scree plot (Figure 1), which shows that three components were above the eigenvalue of 1 and explained 52.67% of the variance. Factor loadings were above the suggested limit of .4 (Pituch & Stevens, 2016) for all of the factors. Therefore, the three-factor resolution was displayed as suitable because it can be supported by the theoretical background. The Cronbach's alpha coefficients (Table 2) were moderate and acceptable. Deleting any questionnaire item would not improve the alpha of the scales.

Scree plot regarding teachers' perceptions
Regarding the validity of the questionnaire items examining teachers' self-reported practice, the KMO was mediocre, .62, and KMO values for the individual items ranged between .38 and .76. The correlations between items were appropriate for PCA as Bartlett's test of sphericity was significant, χ 2 (136) = 336.93, p < .001. All items correlated significantly, .26 ≤ r ≤ .71, p < .001, with at least two other items. Communalities ranged between .29 and .72 and the average was .53. Factor loadings above .4 were considered when determining the factors. Three components had eigenvalues above 2 and the value of the fourth component was slightly higher than 1 (1.16) ( Figure 2); therefore, the three-factor resolution explaining 52.82% of the variance was preferred because it allows the comparison of teachers' perceptions on assessment objectives with the application frequency of these assessment methods. As for the reliability of the scales, all Cronbach's alpha figures ( Table  2) were acceptably high.

Figure 2
Scree plot regarding teachers' self-reported practices

Data Analysis
To answer the three research questions, Statistical Package for Social Sciences (IBM SPSS) V25 was used. First, to characterise teachers' perceptions, descriptive statistical analyses were used. The differences between questionnaire items were also examined using a series of t-tests. Second, the frequency of using TA and AA methods was analysed with descriptive statistics, and the differences between them were revealed by performing t-tests. Finally, the differences between the scales (Table 2) were compared and the relationships between them were explored by calculating the correlation coefficients. Table 3 contains descriptive statistical parameters for the questionnaire items measuring teachers' perceptions about the importance of TA and AA objectives. The statements are presented in descending order of their averages. By performing a series of paired samples t-tests, we examined whether there were significant differences between the means of each two items in this order. Based on all this, the statements can be divided into two groups.

Research Question One
The first group consists of nine statements related to AA. Without exception, there were no significant differences between the means (4.22 ≤ M ≤ 4.67). Typically, most of the teachers involved in our study agreed or strongly agreed with the stated objectives for AA. The use of the lower values of 1 to 3 was not typical. Our sample can be considered homogeneous (0.52 ≤ SD ≤ 0.82). Significant differences (p < .05) in pairs were found only between the means of the first two and the last two statements belonging to this group, most of which are items describing the aims related to AaL.
The second group contains four statements measuring the degree of agreement with objectives related to TA. The separation from the previous group is shown by the significant difference between the means related to the statement "encourage collaborative learning", which refers to AA, and the item "encourage students to recall or recognise facts, ideas, and propositions in life" which is a characteristic element of TA, t(50) = 2.99, p = .004. A significant difference was also found between the means of the last two statements, t(50) = 3.85, p < .001. The heterogeneity of the sample was verified by the larger standard deviations (0.94 ≤ SD ≤ 1.19) compared to the questionnaire items belonging to the previous group (0.52 ≤ SD ≤ 0.82). This means that teachers, who took part in this study, generally agreed with the statements related to TA; but compared to the previous group, the majority of teachers chose values of 2, 3, and 4, and the first scale point in the case of the last statement. Notes. 1 = Strongly Disagree, 2 = Disagree, 3 = Neutral, 4 = Agree, 5 = Strongly Agree. * Mean significantly differs from the previous statement at p < .05. Table 4 contains different assessment methods for which we examined how often the teachers involved in this study stated that they used these methods to obtain information about their self-reported practice. The table shows the application frequency of assessment methods in descending order of the means. Paired samples t-tests were used in pairs to examine the differences between them. We distinguished three major groups.

Research Question Two
In descending order based on the averages, only TA methods are in the first six places. Respondents applied mostly or/and always true/false tests, filling the gaps, and question-and-answer tasks. Matching exercises and sentence completion were sometimes or/and mostly used by these teachers. However, sentence completion was significantly more often used than short-answer tests, t(50) = 2.52, p = .02.
The second group is mixed in terms of its composition because it includes not only traditional but also various AA methods. When examining, in pairs, the application frequency of the TA techniques, chiefly multiple-choice tests, information transfer, and essays, there was no significant difference between multiple-choice tests and information transfer. These methods were sometimes used by about 50% of the respondents, while essays were applied less frequently. The difference between multiple-choice tests and essays was significant, t(50) = 3.00, p = .004. In the cases of AA methods, which can be used to fulfil AaL purposes, the means were still moderate. When determining the application frequency of teacher observation, the sample was heterogeneous (SD = 1.35).
The reason for this heterogeneity may be that this method can be applied both informally and formally and, therefore, there may be large individual differences in the interpretation. Compared to this method, the teachers used peer-assessment significantly less, t(50) = 3.00, p = .004, rarely and sometimes, while there was no significant difference between the frequency of self-and peer-assessment. Among the methods that support learning, projects and presentations were most often used, and there was no significant difference between these two methods in terms of their application frequency.
The third group consists of assessment methods that support learning. The difference between the frequency of application of presentations and reports was significant, t(50) = 5.11, p <.001. However, the frequency of using reports did not differ significantly from that of interviews and portfolios. Teachers, who completed the questionnaire, used conferences the least. As an indicator of this, the difference between this method and portfolios was significant, t(50) = 4.15, p < .001. Conferences had the smallest standard deviation, 0.66, indicating that teachers judged the frequency of its application as the least used method as it was reported to never be used by 73% of the respondents.  Table 5 presents the results of descriptive statistical analyses based on the composite scores of the scales that emerged from the factor analysis of the two dimensions. Higher scores describing teachers' perceptions indicated greater importance of the given assessment objectives, while higher results in the case of teachers' self-reported practice revealed the more frequent use of the given assessment methods.

Research Question Three
As the separate analysis of the individual questionnaire items measuring teachers' perceptions showed (Table  3), Moroccan teachers involved in our study agreed or strongly agreed on the importance of AA objectives. The minimum and maximum values, means, and standard deviations of the created scales describing AA objectives associated with AaL and AfL were also similar, there was no significant difference between the means of these scales. Furthermore, the correlation between these two scales was significant, r = .43, p = .002. The teachers perceived the objectives of AA as more important than that of TA, because a significant difference was found in the case of AaL, t(50) = 9.34, p < .001, and AfL, t(50) = 10.19, p < .001. The respondents were generally neutral when rating the importance of TA objectives and the sample was more heterogeneous compared to the other two scales measuring AA purposes. There was also a significant correlation between perceptions associated with AfL and TA, r = .30, p = .03.
Based on their self-reported practice, the teachers involved in our study used methods associated with AaL more often than those of AfL. The difference between the two scales was significant, t(50) = 4.52, p < .001. The respondents used TA methods more often than those of AA, and the difference was significant compared to AaL, t(50) = 5.71, p < .001, and AfL, t(50) = 13.08, p < .001. The sample was more homogeneous in terms of the frequency of using TA methods compared to AA techniques. Even in the case of self-reported practice, there was a significant correlation between the application frequency of AfL and TA methods, r = .31, p = .03.
When comparing teachers' perceptions with their self-reported practices, no significant difference was found between the means of TA objectives and their associated methods, t(50) = 1.25; p = .22. However, concerning AA, objectives associated with AaL were considered more important by the teachers involved in this research, while the methods of this assessment were much less frequently used. The difference between the means of scales describing the perceptions and self-reported practices regarding AaL was significant, t(50) = 12.93, p < .001.
When comparing the importance of objectives associated with AfL and the application frequency of their methods, a similar tendency was found based on the significant difference between the means of these two scales, t(50) = 18.51, p < .001. However, this degree of difference was somewhat larger compared to the other two scales covering perceptions and methods associated with AaL.

Moroccan EFL High School Teachers' Perceptions about TA and AA
The teachers involved in our study clearly acknowledged the importance of the AA objectives because most of them agreed or strongly agreed with each statement. This was also evidenced by the fact that there was no significant difference, but a close relationship between the scales describing AfL and AaL objectives. Likewise, and also in the Moroccan context, respondents in Ghaicha and Omarkaly's (2018) study expressed their agreement or strong agreement to all questionnaire items related to the effectiveness of AA in helping students demonstrate and apply their competencies in real-life situations, to be more motivated and active during the lesson, and to measure their higher-order thinking skills. Our findings also corroborate the results of Babni's (2019) study where most Moroccan teachers indicated a high degree of agreement with the questionnaire items on AA and perceived it as an imperative approach in English language teaching and learning.
In our survey, teachers also perceived the objectives of AA to be significantly more important than those of TA. This is encouraging, as research evidence strongly demonstrates that AA is advantageous concerning its effective learning outcomes for students (Aksu Ataç, 2012;Al-Mahrooqi & Denman, 2018;Cirit, 2015;Nasab, 2015;Phongsirikul, 2018). The results of our study were also in conformity with studies that compared TA to AA in terms of teachers' perceptions. For example, Babni (2019) found that 74% of English teachers acknowledged that they perceived AA to be much more effective than TA.

Moroccan EFL High School Teachers' Assessment Practices
Based on their self-reported practice, the teachers taking part in this study used TA methods, mainly true/false tests, filling the gaps, question-and-answer tasks, matching exercises, and sentence completion more often than any AA methods. This finding is in harmony with the findings of Ghaicha and Omarkaly's (2018) study, which revealed that teachers have more positive perceptions and greater preferences towards traditional methods rather than alternative ones. The use of other TA techniques, namely multiple-choice tests and information transfer, was mediocre and showed similarities to the application frequency of AA methods associated with AaL. At scale level, the difference between the use of methods linked to AaL and AfL was also significant. The reason for this difference may be that teacher observation was primarily marked as more frequently used than self-and peer-assessment, and secondarily as more similar in terms of the frequency to the two methods of AA associated with AfL (projects and presentations) rather than the remaining methods, chiefly reports, oral interviews, portfolios, and conferences. Additionally, in the study of Ghaicha and Omarkaly (2018), portfolio was rarely used, and the findings of Benzehaf's (2017) study conducted via a questionnaire and interviews demonstrated that only 10% of teachers applied self-and peer-assessment. Therefore, these AA methods have not been utilised in Moroccan EFL high schools. This could be related to teachers' pressure and tendency towards finishing the curriculum that ends with a high-stakes examination (Ghaicha & Oufela, 2021;Ouakrime, 2000). This race against time has become a priority from the perspective of teachers and students regarding the teaching and learning process as well as an obstacle for teachers as they attempt to provide both effective feedback and new learning opportunities (Ouakrime, 2000).

The Relationship Between Moroccan EFL High School Teachers' Perceptions about Assessment and Their Self-Reported Practices
When analysing the relationship between the different scales, we found significant correlations between AfL and TA in both the perceived importance of their objectives and the application frequency of their methods. The reason for these significant correlations can be that both forms of assessment focus on student performance and aim to determine the extent to which learners meet instructional outcomes, although they differ significantly in their emphasis, as the aim of AfL is to enhance the teaching-learning process, while TA refers to the qualification of knowledge acquired at the end of the learning process. In terms of assessment practice, TA methods can also be used to enhance the teaching-learning process and fulfil the purposes of AfL. These are achieved, for example, when students have the opportunity either to check their answers with a key after completing tests requiring short answers or to discuss the results with their peers. AA methods can also be used to assess learning outcomes at the end of the learning process, especially when students, for instance, perform their presentation or share the results of their portfolio at the end of the process and receive grades for these products from their teacher. Based on our survey, we have no information on how the products of AA are incorporated into the final grades or how teachers apply the methods of TA and AA in their classroom practice. However, the significant relationship between objectives and the self-reported application frequency of the methods related to AfL and TA reveals the complex nature of teachers' perceptions and application of assessment. This finding encourages teachers to try various methods of TA and AA rather than focus on one approach (Babni, 2019;Nasab, 2015).
When comparing teachers' perceptions with their self-reported practices, we found that teachers' perceptions regarding TA match their practices because there was no significant difference between the two scales describing the objectives and methods of TA. Therefore, it can be argued that Moroccan teachers believed that they were assessment literate regarding their perceptions of TA and believed they reflected their knowledge in their practices, which was justified by their frequent employment of traditional methods. These are apparently enacted in secondary education through two modes of continuous assessment: three quizzes (15 to 20 minutes) and two summative tests per semester (Ministry of National Education, 2007). This may also indicate Moroccan English teachers' potential for implementing and scoring standardised tests.
As found in previous studies (e.g., Ghaicha & Omarkaly, 2018;Phongsirikul, 2018), our study also revealed that most teachers admitted that they found AA important even though they did not often use it to either have students reflect on their own learning and support AaL or enhance their performance in the learning process and promote AfL. For instance, many of them reported that they appreciate the value of encouraging collaborative learning, but its practice, for example, via peer-assessment, is not emphasised as a frequently used method by Moroccan teachers. This may reveal the teachers' lack of knowledge on how to put the objectives of assessment into practice as well as lack of institutional support. Ghaicha and Oufela (2021) argued that Moroccan school authorities (administrators, school managers, counsellors, and other teachers) do not actually support teachers' use of AA strategies.
A key question that arises from these findings is why such differences in self-reported practices arise between TA and AA methods. In our study, the composition of the sample (Table 1) may have influenced the results. As argued by Hakim (2015), teaching experience can influence the application frequency of AA methods. Unfortunately, we did not have the opportunity to investigate this due to the small sample size. Other possible reasons, as stated by some researchers (Ghaicha & Omarkaly, 2018;Ghaicha & Oufela, 2021;Isik, 2021;Janisch, Liu, & Akrofi, 2007;Nasri et. al. 2010), might be associated with the perceived familiarity with assessment techniques, preoccupation with exam scores, time constraints, limited English proficiency, an increase in teachers' workload, the educational system, classroom size and materials, a lack of assessment-related training courses, and issues of communication and collaboration between different stakeholders (teachers, learners, and administrators). Thus, the related findings should nudge researchers and educators to reconsider teachers' perceptions of assessment and how they should be enacted in practice to resolve the mismatches found in this study.

Limitations and Suggestions for Future Research
This study has some limitations that can be related to a variety of factors. For instance, we examined a few elements of assessment literacy and analysed the characteristics of classroom assessment through teachers' perceptions and self-reported practices, although no information was available about their behaviour. We also lacked a combination of qualitative and quantitative research methods that could have been achieved by conducting interviews with teachers or implementing classroom observations to explain the quantitative results. The study also did not represent the view of all EFL teachers in the Moroccan context and because of the relatively low sample size, we were not able to obtain information about the influencing role of individual differences and classroom environmental factors. For example, no reflection on teachers' perceptions and practices of TA and AA in relation to their teaching experience, age, gender, and the number of students they teach in an average class was utilised. Furthermore, only teachers were included in the study. Despite these limitations, in many cases, we were able to formulate hypotheses about the interpretation of the findings and the reasons for the mismatches found in this study. These problems also indicate possible directions for further research. The developed questionnaire, for which the validity and reliability have been verified, can be applied in further studies and, therefore, can pave the way for researchers to get more detailed information about the classroom assessment beliefs and methods of EFL teachers by involving additional types of research and other relevant stakeholders. Thus, future research is needed to investigate the perceptions and practices related to different CBA approaches that encourage lecturers to rethink and ameliorate their assessment practices by applying varied AA and TA tasks effectively, which allows students to develop and assess their language abilities.

Conclusion
In general, the findings of this study demonstrated that most Moroccan EFL public high school teachers have positive perceptions towards the objectives of TA and AA. They were assessment literate and reflected their understanding of TA objectives into actual use; however, there was a mismatch between what teachers perceive and what they stated that they do in their classroom regarding AA associated with AfL and AaL. Based on these findings, it is deemed necessary for teachers to change their classic teacher-centred assessment to a learningcentred approach. To achieve this, ELT training centres need to adopt solid training that considers the alignment of teachers' understanding of the principles of AA and their actual enactment in the classroom through various types of assessment methods that are effective both during and at the end of the learning process. Schools also need to be encouraged to involve AA activities as a part of their policy and a part of teachers' courses as followup activities. Setting up a system of support that could help teachers conduct AA is also needed in order to overcome some obstacles (e.g., overloaded classes, logistics); thus, classroom management issues should be reconsidered. These recommendations can raise the awareness of English teachers, trainers, and researchers of the numerous AfL and AaL methods that can be implemented in ELT classrooms.