Higher Ed SIG Corner


Jennifer Rockafellow, Purdue University Calumet

Back to the Link front page

Jennifer is currently Assistant Director of the English Language Program at Purdue University Calumet. She joins ITBE with 13 years’ experience in ESL education, including IEP, adult education, ESP, and advising and mentoring of international students. She is currently serving as interim Higher Education Special Interest Group Leader and looks forward to contributing to and learning from colleagues within the organization.
The discussion surrounding placement testing in higher education has been gaining momentum recently, with many institutions looking for low-cost options that offer reliable results and are easy to administer.  At ITBE’s Fall Workshop, the Higher Education Special Interest Group focused on the topic of placement testing and the challenges involved in selecting a new assessment.  Purdue University Calumet recently reviewed two new assessment options (CaMLA and Accuplacer), and this article provides a summary of the advantages and disadvantages of each assessment.
Assessment Delivery Method Skills Assessed Time Price
CaMLA Online Reading, Listening, Grammar, Vocabulary 60 min. $5.95 per test
Accuplacer Online Reading, Listening, Language Usage, Sentence Meaning No time limit (adaptable) $2.20 to 5.50 per section (roughly $10 for all of the skills listed)



Developed by the University of Cambridge and the University of Michigan, this test is available as a paper-based test as well as an internet-based test (called the EPT).  The version of the test being reviewed here is the EPT test.

Positives:  The test is administered online and only requires 60 minutes to administer.  There are several versions of the test available so that a new test can be taken each semester to alleviate concerns over leaked or compromised test questions.  The skills tested closely aligned with the placement test that is currently used at Purdue University Calumet. Although setting up administration online was at times challenging, customer support e-mails were returned in a timely manner with effective troubleshooting information.

Negatives:  One of the primary concerns with the test was the listening comprehension portion.  We experienced buffering issues with the audio, resulting in the first word or two being eliminated from the prompt.  Although this may have been an issue stemming from poor connectivity in the testing room, if this were to occur on testing day, the results would be invalid.  The listening comprehension passages and prompts were only stated once with no option to repeat.  Additionally, the committee thought the items in the vocabulary section of the test were difficult.  While high-intermediate and advanced students could have completed the test with accurate results, we felt the vocabulary items were too advanced for low-intermediate and beginner students and the test would be overwhelmingly difficult.  Navigation within the test was also problematic, with prompts appearing to instruct the student how to continue.  This may be an issue for low-level students who are unable to read and understand the prompts during a self-guided test.  Lastly, CaMLA states that a scoring report is generated for each student and sent back to the administrator, though we never received a report following the scoring of our pilot exam. 


Positives:  Unlinke CaMLA, the test is adaptive, ideally resulting in a more tailored assessment.  The committee felt that the listening comprehension portion of the test was useful.  Students can replay the listening prompts and the topics were relevant for an IEP.  Time limits for each section can be determined by the institution.  Placement testing for domestic students at Purdue University Calumet is already done using Accuplcer, so there is a high degree of institutional support.  Easily accessible validity studies were also available.

Negatives:  Each segment of the exam is purchased separately, meaning that a test consisting of four skill areas (listening, reading, language usage, sentence meaning) is about $10/test, making it less cost effective than the CaMLA.  The committee also had some problems recognizing how the test adapted to an incorrect answer.  For example, when answering a question incorrectly, the subsequent question sometimes appeared less difficult, though this was not always the case.  Sections of the vocabulary test were also very advanced (testing comprehension of advanced forms such as phrasal verbs, for example).  Lastly, the score that was immediately generated was a numeric score, though it was unclear what the maximum score was, which percentile the score was in, how the score could be plotted against the CEFR levels, etc.  The institution would need to develop its own conversion chart and appropriate level cutoffs in order to adopt this assessment.

While both assessments had notable advantages as well as disadvantages, overall the committee at Purdue University Calumet determined that neither of the assessments met the program’s needs.  The English Language Program has subsequently decided to cancel any planned student pilots and resume the search in spring 2016 beginning with the Oxford Online Placement test.


Back to the Link front page

ITBE Link - Winter 2016 - Volume 43 Number 4