The OET Centre works with the Language Testing Research Centre at the University of Melbourne to ensure the reliability of test results and analysis. Each of the four sub-tests is assessed in a specific way (summarised below).

Boards and councils can access results via the OET website. Registration authorities are encouraged to register for secure access to the site to check individual results; access to the online verification system is controlled by OET Centre management. In order to gain access, managers working for health profession Assessing/Registration Authorities can apply for individual staff members to have secure access to the relevant function of the OET website. At present a number of staff of the DIAC Skilled Processing Centre (Adelaide) and staff of 13 Australian & New Zealand Health Profession bodies are registered to do so.

How are OET results reported?

Candidates receive results for the four sub-tests that make up the OET as one of five grades – A to E.

A is the highest grade, E is the lowest.

There are descriptions for each grade on the printed Statement of Results candidates receive:



There is no overall grade; candidates receive a separate grade for each sub-test he/she takes.

Most boards and councils that recognise the OET require candidates to have at least a B grade in each of the four sub-tests. However, candidates should check with the boards and councils that regulate their profession to confirm what is currently required. The OET Centre cannot take any responsibility for providing or checking this information for candidates.

Candidates may receive different grades for your performances in different sub-tests. There are many aspects to being able to use and communicate in a language effectively and the OET tests many of these across the four sub-tests. Test design separates test materials into the four macro-skills (Listening, Reading, Writing and Speaking) for practical reasons and to acknowledge that these involve different sub-skills and strategies.

The length of time the results are recognised as valid is decided by the authorities (boards and councils; government departments) that recognise the OET. This period is usually two years, but candidates should check with the boards and councils that regulate his/her profession to confirm.

Each sub-test has its own two-year validity. A set of results for the four sub-tests is valid for as long as the oldest satisfactory sub-test result remains valid.

It is the policy of most boards and councils that test takers must achieve the requisite grades for each sub-test of the OET in one sitting. This is not a policy as set by the OET Centre.

Answer sheets for Part A are scored by trained markers in Melbourne who work with the aid of a comprehensive Marking Key. Score sheets for Part B (multiple-choice responses) are computer-scanned by the University of New South Wales.

Part A Reading scores

Each section of the new format Reading sub-test is weighted equally. That is, Part A is worth 33.3% and each text in Part B is also worth 33.3% of the total score.

Why is the Reading sub-test weighted equally when there are more questions (items) in Part A?

When candidates sit the reading test, they will get a raw score on Part A (between 25 and 35 depending on the number of items on that particular test) and a raw score on Part B (between 16 and 20 depending on the number of items on that particular test). Scores will then be converted to a total score for the reading test in which a candidate' score for Part A counts for 33.33% (one-third) of the total score, and the score for Part B counts for 66.66% (two-thirds) of the total score. A candidate's total score for the reading test will then be transformed into a grade (A – E).

This type of weighted score conversion is very common. For example, in the listening sub-test, although one Part A might have more available marks than Part B, a candidate's score on each part is converted so that it counts for exactly 50% of his/her total score for the whole listening sub-test.

The reason why there are different weightings for each part of the reading test is to reflect the relative importance of the skills being measured in each part in the overall estimation of a candidate's reading ability. Part A tests a relatively narrow set of reading skills – skimming and scanning – whereas Part B tests a range of different reading skills through different types of multiple choice questions.

Do candidates need 65% for each Reading section individually?

Part A and Part B scores are converted to a percentage (reflecting relative weighting) that is then converted to a total score. Data from candidates' total score across both sections are used to calculate band-scores for each administration. The score is a combination of total weighted scores for Parts A and B; candidates do not therefore necessarily need to achieve the exact score that equates to a ‘B’ (for a given administration) in both sections. Conceivably, candidates may score slightly lower for one section but still maintain an overall higher level if scores for the other section are higher.

The Listening sub-test is marked in Melbourne. The Listening sub-test is assessed against a detailed marking guide prepared by the test designers. Problematic scripts are dealt with as a group by an experienced marker and all critical borderline scripts are double-marked. Before each administration, test developers, the OET Assessment Manager and markers for the Listening sub-test conduct a formal meeting to finalise the marking guide, test audio and test-paper and to moderate their initial marking. A number of ‘live’ control scripts are utilised for the purpose of ensuring standardised marking among the marker cohort prior to the commencement of the assessment period.

Each test taker’s writing paper and recorded speaking performance is marked by two independent raters in Melbourne. Sets of raw scores provided by each rater are sent to the Language Testing Research Centre where the test analysis program FACETS (Linacre, 2010) is used to analyse the score data. FACETS relates task difficulty, test taker ability, and rater harshness/leniency to each other. This allows it to adjust test takers’ scores for the effect of an overly harsh or lenient rater. FACETS produces a fair-score that takes rater harshness and task difficulty into account. It also identifies raters who are overly harsh or lenient, or who rate inconsistently or too conservatively. These raters can then be retrained.