September 2009 home | PDF |

Title
An Investigation into the Validity of Conversational C-Test as a Measure of Oral Abilities

Authors
Purya Baghaei (Ph.D.)
Islamic Azad University, Mashad, Iran
Mohammad Monshi Tousi (Ph.D.)
Islamic Azad University, Mashad, Iran
Ali Akbar Boori (M.A.)
 Islamic Azad University, Mashad, Iran

Bio Data
Purya Baghaei, assistant professor in the English Department of Islamic Azad University, Mashad, Iran. His main research interests are language testing and the application of Rasch modelling in social science research.
M.T. Monshi Toussi, assistant professor in the English Department of Islamic Azad University, Mashad, Iran. His main fields of interest include translation, and research on EQ and affective factors in language teaching and learning.

Ali Akbar Boori, lecturer in the English Department of Islamic Azad University, Mashhad, Iran. His main research interests are language testing and teaching.

Abstract
Alderson (2002) argues that C-Test is not an appropriate measure of language proficiency or even placement. His rationale is that there are no aural and spoken elements involved in C-Test taking, so they do not tap oral-aural skills. Nevertheless, there is evidence from previous research which asserts that text type can affect what a C-Test measures (Baghaei, 2008; Daller & Grotjahn, 1999; Sigott, 2004). The major goal of this project is to take advantage of this finding and test whether C-Tests constructed out of spoken discourse (dialogues, interviews, etc.) will result in the prediction of listening skill of the test-takers. In this study, two C-Test passages were constructed out of spoken language tests, and two C-Test passages out of written language texts. The four passages were then put in a C-Test battery and were administered along with multiple choice tests of reading comprehension and listening comprehension to a sample of undergraduate English students. Regression analysis indicates that conversational C-Test passages are better measures of reading, listening and composite scores of reading and listening than written C-Test passages.

Key words: conversational C-Tests, written C-Tests, cloze test, correlation,   regression