Authentic, Performance-Based Assessment in ESL/EFL Reading Instruction
Title
Authentic, Performance-Based Assessment in ESL/EFL Reading Instruction
Authentic, Performance-Based Assessment in ESL/EFL Reading Instruction
Keywords: No Keyword
Authors
Richard Lynch
Richard Lynch
Bio Data
Richard Lynch teaches in the graduate program in multimedia-assisted language learning at Woosong University, Korea. He has a Certificate in TESL from Concordia University, Montreal, an MA/TEFL from Southern Illinois University, and a Ph.D. in educational psychology/instructional technology from the University of Southern California. He has taught ESL/EFL in Canada, the U.S., Papua New Guinea, Japan, Thailand, and Korea. His research interests include learner performance and motivation in diverse distance learning contexts including second/foreign language learning.
Abstract
In recent years there has been a reaction against standardized objective testing and a movement toward authentic, performance-based testing in a variety of learning domains, including ESL/EFL reading comprehension. A number of measurement researchers have investigated and developed comprehensive test validity criteria which should apply to all testing, traditional or authentic. The advantage of performance-based testing resides in its potential to engender and sustain positive washback on the teaching and learning process. Such positive impact on the instructional process is not, however, a sine qua non of performance-based testing. It can only derive from a comprehensively valid interaction between the nature of the instruction preceding evaluation and the actual performances being assessed. Validity must be grounded in a range of interrelated factors which can all be subsumed under a general notion of construct validity. In order to maximize the validity of performance-based assessments both test designers and teachers need to be aware of these factors and their interaction. Construct validity and its interrelated aspects are discussed and applied to the authentic, performance-based testing of reading comprehension. A reading exercise is presented and discussed as a representative instructional model that can be used to inform valid performance-based reading comprehension tests.
In recent years there has been a reaction against standardized objective testing and a movement toward authentic, performance-based testing in a variety of learning domains, including ESL/EFL reading comprehension. A number of measurement researchers have investigated and developed comprehensive test validity criteria which should apply to all testing, traditional or authentic. The advantage of performance-based testing resides in its potential to engender and sustain positive washback on the teaching and learning process. Such positive impact on the instructional process is not, however, a sine qua non of performance-based testing. It can only derive from a comprehensively valid interaction between the nature of the instruction preceding evaluation and the actual performances being assessed. Validity must be grounded in a range of interrelated factors which can all be subsumed under a general notion of construct validity. In order to maximize the validity of performance-based assessments both test designers and teachers need to be aware of these factors and their interaction. Construct validity and its interrelated aspects are discussed and applied to the authentic, performance-based testing of reading comprehension. A reading exercise is presented and discussed as a representative instructional model that can be used to inform valid performance-based reading comprehension tests.
[private]
[/private]
Category: Main Editions
