Placement Tests and How to Improve Them

This is the summary of the #ELTchat on 7 September 2011 9:00 pm BST moderated by @Shaunwilden and @Marisa_C  and written by Natasa Nikpalj-Juraic who tweets as @Notyetlanguage.

Introduction

Last #ELTchat focused on Placement Tests as we believe that it’s not easy to design and implement them. Obviously, at the the time of enrollment, we need a reliable and not time-consuming tool to place students into, as much as it is possible,  homogeneous groups. If we fail, poor level placement might jeopardise our students’ learning objectives and what’s even worse, make them feel uncomfortable in class.

Definition and purpose of placement tests

Everybody agreed on these definitons:
@TyKendal: tests to ascertain a ss level in order to stream them accordingly
@PrettyButWise: A test to determine level of students so you can place them in the right class or design the right course for them
@padget for me placement tests give me a snapshot of abilities so I know where to start to make content accessible

@ELTExperiences stated their two purposes:

1. To check grammar/vocabulary knowledge
2. To place SS into appropriate level

Drawbacks

It seems that we all agreed that placement tests should test language skills and communicative competence, not specific knowledge. The main drawback of most placement tests is its concentration on decontextualised grammar (esp. verbs and tenses) and vocabulary.
@TyKendall: scary thing I noticed is that most online place/tests are purely grammar based and specifically VERB FORMS is this all language is?
@Marisa_C: SOME vocabulary but no skills sections
@ELTExperiences: Placement test for YLs at local school assesses grammar knowledge and places students into levels. But too much focus on grammar
@TyKendall: skills are almost completely neglected in most placement tests
@ELTExperiences: a poor placement test is checking grammar skills and not really focused on communicative competence
@PrettyButWise: i think most of online placement tests check grammar skills not communicative competence
@jorech: Too often, placement tests are too short , which breeds preconceived expectations
@ELTExperience: If a school promotes ‘communicative competence’ why are we testing grammar and not speaking, pronunciation, etc.?

Written, oral or both?

We also concluded that most placement tests are mostly done in written format in schools and are mostly multiple choice tests, but the good news is that sometimes it includes a piece of writing and an interview. We agreed that productive skills should not be neglected.
@ayatawel: It’s mostly done in written format in schools
@PrettyButWise: here in Kuwait they are written tests

Ways to improve them and examples of good practice

There was an interesting comment by @ELTExperiences who thinks that needs analysis should come before any placement. It could definitely help us place students, but also design syllabi. @sandymillin tends to do needs analysis, since placement test is only seen by management and administration.

@Marisa_C considers that teachers should be involved in placement testing and allowed to see results. In some places, like @dreadnought001‘s university , the tests are reasonably comprehensive – they include grammar, listening and speaking.
To overcome the drawbacks of short, multiple-choice tests, it’s also possible to use the final exam as the placement test, e.g. when you are testing students for placing in Exam Classes, you could use the exam itself (e.g. FCE) i.e. past papers as a placement instrument (@Marisa_C, @TyKendall)

In connection with the importance of the right placement of students some participants said they’ve never been fond of mixed -level classes. We also shared some of our own experiences of taking placements test as students of other languages and most participants said that they were placed into the correct class. This means that in spite of their disadvantages placement tests often fulfil their purpose.
It’s also important to help students relax before taking a placement test; otherwise the results won’t be accurate. Furthermore we can use placement tests results to help learners establish goals and plan our syllabi.
@eyespeakbrasil: The beginning for a good placement test, is to design the assessment parameters first
@hartle: We often use Oxford Placement tests at uni, but then do activities in first lessons to further monitor placement
@sandymillin I’ve done a placement test for French about 10 years ago and it placed me into the correct class
@eyespeakbrasil: Placement tests need to be integrated into the needs analysis in my opinion – one in the same
@PrettyButWise: Sometimes students are stressed when they first arrive so you must take time help them relax otherwise test won’t be accurate
@ShellTerrell: I use placemt tests results to help learners establish goals! They plan out what they will do to improve those skills
@sandymillin: I’ve been fortunate, all my institutes have tested listening, speaking, reading and writing
@eyespeakbrasil: I think a spoken placement test, that is recorded, compared to a teacher, and marked would be interesting
@Marisa_C: ask students to work together in groups to see how they would improve the placement test

In a perfect world, placement tests would test all language skills as well as communicative competence and be integrated into needs analysis. However, even in a perfect world, we would always have to spice up placement and testing with our personal judgement and professional knowledge and skills since after all, level is in any case a very fluid thing as @hartle rightly said. Even with placement tests language learners will each have various strengths and weaknesses (@ShellTerrell).

We would like to thank @Shaunwilden and @Marisa_C for moderating yet another motivating ELTchat.