---

Local independence of test items is an assumption in all Item Response Theory (IRT) models. That is, the items in a test should not be related to each other. Sharing a common passage, which is prevalent in reading comprehension tests, cloze tests and C-Tests, can be a potential source of local item dependence (LID). It is argued in the literature that LID results in biased parameter estimation and affects the unidimensionality of the test. In this study the effects of the violation of the local independence assumption on the person measures in a C-Test are studied. A C-Test battery comprising four passages, each containing 25 blanks, was analysed twice. Firstly, each gap was treated as an independent item and Rasch’s (1960) dichotomous model was employed. In the second analysis, each passage was treated as a super item and Andrich’s (1978) rating scale model was used. For each person, two ability measures were estimated, one on the basis of the dichotomous analysis and one on the basis of the polytomous analysis. The differences between the two measures, after being brought onto the same scale, are compared and the implications are discussed.

---

This article investigated and compared the consistency of self and peer assessments as alternatives for teacher assessment.  Thirty sophomores majoring in TEFL were asked to assess their classmates’ as well as their own speaking ability in a conversation class. They were taught how to do this using a rating scale of speaking. They did the rating twice during the term the first rating was carried out during the 8^th and 9^th weeks and the second rating at the end of the term (weeks 15 and 16). The results of the study indicated that self and peer assessments were not significantly related at the end of the term and only loosely, though significantly, related in the middle of the term. Both self and peer assessments indicated consistency over time, however peer assessment enjoyed a higher consistency.

---

The purpose of the present study was to develop appropriate scoring scales for each of the defined stages of the writing process, and also to determine to what extent these scoring scales can reliably and validly assess the performances of EFL learners in an academic writing task. Two hundred and two students’ writing samples were collected after a step-by-step process oriented essay writing instruction. Four stages of writing process – generating ideas (brainstorming), outlining (structuring), drafting, and editing – were operationally defined. Each collected writing sample included student writers’ scripts produced in each stage of the writing process. Through a detailed analysis of the collected writing samples by three raters, the features which highlighted the strong or weak points in the student writers’ samples were identified, and then the student writers’ scripts were categorized into four levels of performance. Then, descriptive statements were made for each identified feature to represent the specified level of performance. These descriptive statements, or descriptors, formed rating scales for each stage of the writing process. Finally, four rating sub-scales, namely brainstorming, outlining, drafting, and editing were designed for the corresponding stages of the writing process. Subsequently, the designed rating scales were used by the three raters to rate the 202 collected writing samples. The scores thus obtained were put to statistical analyses. The high inter-rater reliability estimate (0.895) indicated that the rating scales could produce consistent results. The Analysis of Variance (ANOVA) indicated that there was no significant difference among the ratings created by the three raters. Factor analysis suggested that at least three constructs, –language knowledge, planning ability, and idea creation ability – could possibly underlie the variables measured by the rating scale.  

---