Self-evaluations and the acquisition of pronunciation of French as an L2
Abstract: This paper proposes to look at self-evaluations as a tool in the acquisition of French pronunciation as an L2. The purpose of this paper is, first, to analyze the data from a beta-pilot test of the instruments created by triangulating information gathered from item analysis on the two instruments created (the sentences to record and the self-evaluations) to check the quality of the items. This paper will also assess whether the use of self-evaluations by college-level learners of French enrolled in a phonetics course, will improve their pronunciation over the course of a semester. It will also evaluate whether, and to what extent self-evaluations items completed by the learners are reliable and valid. The students have been recruited from the French phonetics course, and then divided into a treatment and control group. They submitted recordings of text and sentences provided by the researcher before and after a specific phonetic lesson and submitted self-evaluations before and after completing the recordings (treatment group only).
Self-evaluations in the acquisition of pronunciation of French as an L2
Since the introduction in the 1970s of the Communicative Language Teaching Method (CLTM), teaching and learning pronunciation has been unpopular. Before the 1970s, the audio-lingual methods focused intensely on pronunciation, it then fell into disuse. According to Thomson & Derwing (2014), “pronunciation fell out of vogue” (p.9). Most researchers came to ignore pronunciation learning and teaching. However, in the past decade pronunciation instruction has progressively become significant again (Thomson & Derwing, 2014). Researchers have shown that pronunciation instruction can be beneficial to learners of foreign languages (Birdsong, 2007; Couper, 2006; Derwing & Munro, 2015; Grant, 2015; Harding, 2013; Lee, Jang, & Plonsky, 2014; Levis, 2005; Saito, 2011). Pronunciation teaching is now moving towards comprehensibility and intelligibility rather than native-likeness, as the latter is near unattainable (Birdsong, 2007). There is, though, one problem when it comes to teaching pronunciation: the lack of efficient, valid, and reliable methods to assess learners’ pronunciation, which Saito (2011) argues, is one key element to the success of pronunciation teaching. Finding valid and reliable methods to assess students\’ pronunciation is key to improving pronunciation instruction and acquisition.
One potentially useful method to assess learners’ pronunciation is self-assessment, since, “if students can appraise their own performance accurately [emphasis added] enough, they will be able to make teachers aware of their individual learning needs” (Blanche & Merino, 1989, p.313). Much of research on self-assessment has been conducted on reliability and validity (Dlaska, 2008), and Fraser (2000) warned about the need for a reliable methodology behind self-assessment in teaching pronunciation. The impact of such a method on learners’ pronunciation, whether negative or positive, needs further investigation. Because students are the center-part in their own learning and need to be more proactive (Salimi, Asghar Kargar, & Zareian, 2014), it is also necessary to examine students’ awareness of their own learning progress.
This paper proposes to look at self-evaluations as a tool in the acquisition of French pronunciation as an L2 and test self-evaluations’ reliability and validity by doing an item analysis. It will analyze the data from a beta-pilot test of the two instruments created (sentences to record and self-evaluations). This paper will also assess whether the use of self-evaluations by college-level learners of French enrolled in a phonetics course, will improve their pronunciation over the course of a semester. The aspects of French pronunciation studied here, have been carefully selected because they represent segmental (/y/ vs /u/) and segmental/suprasegmental (schwa) features in French, and they are particularly critical for comprehensibility. This paper seeks to answer the following questions:
RQ1: Are the instruments created valid and reliable tools to assess pronunciation?
RQ2: Is there a difference between control and treatment groups overall and on both aspects: segmental and segmental/suprasegmental?
RQ3: To what extent do the students’ self-evaluations compare with evaluations by the expert rater overall and on both aspects: segmental and segmental/suprasegmental?
Self-assessment is described as a type of formative assessment. Formative assessment is student-centered and differs from summative assessment which is teacher-centered. Formative assessment occurs during the learning process, and not at the end of the learning period. According to Fulcher (2016), formative assessment’s purpose is to “inform and improve learning, rather than simply to assess whether the learners have mastered the learning objectives”. Summative assessments are made of criteria set by the instructor while formative assessments are types of assessment that provide feedback on learner’s performance in order to improve this performance.
In an article published in 2006, Nicol & Macfarlane-Dick argue that “formative assessment and feedback should be used to empower students as self-regulated learners” (p.199). Ross (2005) praises the use of formative assessments and studies the validity of formative assessment comparing it to summative assessment. He concludes his article by saying that formative assessment has a positive impact on listening. Self-assessments are also involved in self-regulated or self-monitored learning. Dlaska & Krekeler (2008) support this idea. Their article explores, however, the reasons why advanced L2 learners (German) find self-assessment of pronunciation difficult, focuses on reliability: the agreement between learners and professional raters on the self-assessments. They agree that “it is one of the main aims of self-assessment that learners are being equipped to monitor their own learning process competently to decide for themselves which areas of pronunciation they wish to improve” (p.507).
In a later study, Butler & Lee (2010) propose two aspects of self-assessment: as a measuring tool and as a tool to facilitate learning, and explore how the validity of such tool is influenced by many factors: “(1) the domain or skill being assessed; (2) students’ individual characteristics; and (3) the ways in which questions and items are formulated and delivered.” (p.7).
In an article targeting the Spanish language classroom, Geeslin (2003) explains why students and teachers should use self-assessment in the classroom, and how to use self-assessment. The author clearly links self-assessment to formative assessments and to the communicative language teaching approach. She states “the incorporation of student self-assessment creates a constructive dialogue between student and instructor and increases the degree to which students are aware of instructional goals for individual assignments (p.857).
In their study published in 2014, Salimi, Asghar Kargar, & Zareian, analyzed learners of English as a foreign language (EFL) and self-assessment in order to make the learners more autonomous and to help them take responsibility in evaluating their own performance. The authors investigated the validity and reliability of self-assessment of pronunciation. It is stated that training the learners to evaluate their performance is essential to reliability. The results of the study are positive towards self-assessment.
Lappin-Fortin & Rye (2014) investigated the relationship between self-evaluations completed by students and evaluations are done by outside evaluators, or expert raters, in a French pronunciation course. In this study, the researchers focused on general aspects of French pronunciation. They found that there was a significant correlation between the students’ self-evaluations and the experts’ evaluations. Students ranked themselves in a similar way to that of the experts. Additionally, the students’ pronunciation improved over the instructional period. However, this study did not have control and treatment groups, which poses a problem to the pronunciation improvement results, and while having qualitative data in the form of open-ended questions, the researchers did not make use of this data and only presented quantitative results.
Préfontaine (2013) states in her study that, some studies support self-assessment, while others question the validity of self-evaluation methods. She found “a lack of correlation between students’ self-assessments and teacher ratings” (p.329). Préfontaine also states that one of the limitations of her study is the lack of qualitative data. Additionally, Ross (1998), in a meta-analysis of studies investigating self-assessment in L2 learning, asserts that one of the main limitations to self-assessment studies is the lack of qualitative-oriented studies.
This study will contribute to investigating validity and reliability of self-evaluations, in an effort to alleviate these inconclusive results due to lack of theory and methodological dead ends.
Self-evaluations, in this study, targeted two specific features of French pronunciation: segmental /y/ vs /u/ and segmental/suprasegmental “silent e” (or schwa). These two features have been chosen as variables because they are critical to learners’ intelligibility and comprehensibility. Munro & Derwing’s (1995) Intelligibility/ Comprehensibility principle asserts that intelligibility is the extent to which a given utterance is understood by the learner, while comprehensibility is the learners’ perception of how well they understand an utterance.
Participants and sampling
The study took place at the University of Illinois at Urbana-Champaign. L2 students in a French phonetics course were recruited to participate in the project via an in-class introduction and follow-up emails. The tasks were integrated into the structure of the French phonetics course, and all students in the course were expected to complete these tasks. The instructor of the course was not a member of the research team.
The Fall 2016 French phonetics course had two sections, one section was the control group (G1) while the other the treatment group (G2). The groups were assigned randomly to one or the other section.
Recruitment took place at the beginning of a class period after permission had been granted by the instructor. The researcher then explained the goals of the study and distributed individual sign-up sheet to preserve the anonymity of the participants. Any student who wished to participate was welcome. The researcher hoped to recruit at least 15 participants in each section of the French phonetics course to meet the requirement for representativeness, but due to lack of enrollment, there were only 7 participants per group. The qualitative data from the participants provided rich enough data to obtain a credible picture and ensure saturation. Thus the requirements for the representativeness/saturation trade-off was met.
Both groups received the same instruction in French phonetics and pronunciation. The phonetics course was held every Monday, Wednesday, and Friday for fifty minutes. Fridays were dedicated to lab work, while Mondays and Wednesdays were lectures. At the University of Illinois, French pronunciation is taught following an explicit methodology. Each phonological feature is explained in detail according to the manner of pronunciation: tongue position, jaw position, lips, etc.
Before the first phonological feature was taught, the participants completed the pre-test (Time 1). The post-test (Time 2) was completed after the instruction of the features. Both pre-test and post-test included two types of reading/recording exercises: a short text and short sentences (created by the researcher), targeting specific phonological features of French: /y/ vs. /u/, or the “silent e” (or schwa). While reading the texts and sentences, each participant was required to record themselves at Time 1 and at Time 2. The recordings took place in the phonetics laboratory at the University of Illinois, where participants can be monitored. The researcher asked the students to record themselves only once to control for repeated recordings, which may allow the students to modify their pronunciation.
...(download the rest of the essay above)