Open Collections

Entrance testing at a private English language institute : toward more accurate placements

University of British Columbia

At a large school for adult international ESL students, the placement test, testing process, and placement of students were perceived to be problematic. Three research questions were asked: 1. What are the historical and institutional influences and limitations that have shaped the institute's placement testing process? 2. On what basis do those concerned think that the test is flawed, and how widespread is the problem? 3. Based on the findings of the previous research question as well as test analysis and participant observation, where might problems with the institute's placement testing and the resulting placements lie? A qualitative design known as evaluative case study was used to describe and analyze the placement test and testing process. To answer question #1, document analysis was used, along with participant observation. To answer question #2, a questionnaire was distributed to all instructors and two administrators, and both were interviewed informally and formally. To answer question #3, for Part One (grammar, vocabulary, reading) of the placement test, 571 answer sheets were analyzed, and two sets of data were correlated using the Pearson product moment correlation coefficient. For Part Two (guided writing), 121 answer sheets were analyzed for completion and trends that might emerge. For Part Three (interview), the aforementioned questionnaire and interviews were used, and test-takers and teacher-interviewers were observed on five registration days. Findings related to question #1 indicated that historical and institutional influences and limitations have shaped the placement testing process, with expansion of the school playing an important role. Those related to question #2 indicated that the faculty and administration felt that the test was flawed because of their belief that components of the placement test were of limited usefulness and that the entire placement process was only somewhat useful to them. Additionally, interview and questionnaire data revealed that the problem was widespread. Those related to question #3 indicated that there were problems with each of the three parts of the test. Specifically, the correlation between level as assigned by Part One and level as assigned by interviewer, while high (r=0.84), was nonetheless not perfect. Furthermore, the correlation could not take into account that the difference between levels at VELI were not equal. Part Two, the writing test, did not distinguish among levels in its first section, and in its second section, a disparity existed between the directions and a strip of cartoon-like drawings that caused confusion among the test-takers. For Part Three, teacher-interviewers indicated that both a lack of formal training in interviewing and varying competencies of interviewers might contribute to misplacements. Implications included guidelines both for the institute to follow to contribute to a more accurate placement test and process, and for other language institutes or programs to use in the creation of useful, accurate placement tests.

Thesis/Dissertation

application/pdf

2009-06-16

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/9231

Language and Literacy Education

University of British Columbia

UBCV

DSpace