Semantic Cues, Grammatical Cues, and External Cues
This study examines two facets of the word class problem: (1) the kinds of linguistic cues that learners attend to when making word class distinctions, and (2) the source of word class errors. In this study, cue is roughly synonymous with the term knowledge source used by Huckin and Bloch (1993) in reference to grammatical, morphological, and other types of information used by learners when inferring word meaning from context. The coding scheme was based partially on previous studies of lexical inferencing (Nassaji, 2003) and incidental vocabulary learning (Paribakht & Wesche, 1999). However, some of the knowledge sources identified in those studies are not relevant to the problem of word class distinctions. Consequently, the coding scheme is a partial adaptation of categories from previous studies, but also derives from the data itself (i.e. data-driven).
A total of 10 cues were identified and subsequently used to code the data. These fell into three main groups: semantic cues, grammatical cues, and external cues. Learners use semantic cues when they search for the meaning or LI equivalent of the target words; grammatical cues encompass a variety of sentence-level cues, such as syntax, derivations, and inflections; external cues include looking beyond the sentence in order to arrive an answer (e.g. perceived word frequency or familiarity with particular forms). The taxonomy of cues is presented with relevant examples in the results section. A secondary analysis was performed on the items that were answered incorrectly on the FCT in order to determine the source of learners’ word class confusions (research question 3). As noted previously, an incorrect answer on the FCT may be the result of the learner’s inability to identify the syntactic slot as requiring a particular type of word. Henceforth, this problem stemming from incomplete syntactic knowledge is referred to as problem 1. A distinct problem involves learners’ misclassification of one or both of the target lexical items. This misclassification is referred to as problem 2. At times, there is evidence for both problems occurring simultaneously, which was coded as ‘combination’ error.
There are three issues worthy of note regarding the coding of the think-aloud protocols. First, given the potential problem of veridicality associated with retrospective reports, the data analysis was based primarily on the concurrent think-alouds. The retrospective reports were used only to verify certain details that may have been omitted during the concurrent think-alouds. Likewise, the retrospective reports are more metalinguistic in nature because participants were given the opportunity to elaborate on their responses, which often lead to explanations and reasoning. Second, participants often used more than one cue per item, while for some items it was impossible to determine what cues, if any, the participant was attending to. For example, if a participant read the sentence aloud and immediately circled one of the forms to complete it without any verbalisation of his/her thought process; this instance was coded as N/I, or not identifiable. Finally, the items on the FCT for which a participant marked ‘neither’ were discarded from the think-aloud analysis.
This step was deemed necessary because the overwhelming majority of ‘neither’ responses reflected the participants’ decision to ignore an item entirely. To ensure the reliability of coding the data for cues, a second rater coded 2/3 of the data set. The second rater was an experienced language teacher who had prior experience with gathering and analysing think-aloud protocols. Inter-rater agreement for that portion of the data was 88%, and subsequent discussion of the discrepancies led to 100% agreement.
With years of experience in making delicate Cheap Louis Vuitton Handbags, we are waiting to provide the crowning touch to your style. Just feel free to shop Louis Vuitton solar PM shoulderbag M93125 in red here.
Processing your request, Please wait....
