• Background
• Test Developers
• Test Publishers
• Test Users
• Technological Issues
• Quality Issues
• Control Issues
• Security/Privacy Issues
• Official Translations
• Acknowledgements
• References
• Appendix
• Glossary

International Test Commission

International Guidelines on Computer-Based and Internet Delivered Testing

Attend to quality issues in CBT and Internet testing

1. Assess the appropriateness of the content and technical adequacy of CBT/Internet testing relative to alternative testing methods for each client. Inform test-taker s of the purpose of the testing so they are able to make an informed decision on the appropriateness of the test for their situation.
2. Have adequate knowledge of the CBT/Internet test and its modes of operation. When required, attend appropriate training events and read and have knowledge of relevant training materials.
3. Follow best practices in the use of CBT/Internet testing and, where appropriate, create ‘best practice’ testing policies.
4. Verify that test-takers know how to interact with an Internet testing system (e.g., basic browser operation, use of access passwords).
5. Maintain and regularly update knowledge about CBT/Internet testing, including pertinent changes in legislation and policy.
6. Adhere to legal, professional, and ethical mandates related to CBT/Internet testing.
7. Inform test-taker s of the limitations of the Internet test in terms of the professional relationship expected from this medium.
8. For Internet testing, provide a contact point (e.g., email or phone) for those who do not understand the purpose of the test.

1. Ensure that documentation of the appropriate psychometric evidence is supplied with the CBT/Internet test.
2. Ensure that current psychometric standards (test reliability, validity etc.) have been met even though the way in which the tests are developed and delivered may differ.
3. Be able to distinguish between tests with and without documented psychometric properties. Those with documented evidence ensure that the evidence is appropriate for the intended use of the test.
4. For Internet testing, use only those websites supported by publishers who offer validated psychometric tests.
5. Check that the CBT/Internet test does not require knowledge, skills or abilities that are irrelevant to the construct being assessed.
6. Where appropriate, review and understand the documentation that describes how the CBT/Internet test uses algorithms for item generation, selection, or test construction, for controlling the order of testing, and the model underlying the development of the test.
7. When necessary, access appropriate training to ensure continuing professional development.
8. Document information provided about changes to test items or parameters and their impact on the test properties.

1. Confirm that the evidence regarding the equivalence of the CBT/Internet test to the manual version is sufficient.
2. If norms are based on manual versions of the test, confirm that evidence has been obtained to show equivalence of test means and SDs across versions and for appropriate subpopulations.
3. Verify that the technical features of the CBT/Internet test (e.g., test-taker control and item presentation) allow the results from the CBT/Internet test to be equivalent to the manual version.
4. Only use the test in those modes of administration for which it has been designed (e.g., do not use a test in an unproctored mode when it is specified for use only in proctored modes).

1. Review and understand the rules underlying the scoring of the CBT/Internet test.
2. Inform test-takers, when appropriate, about how scores are generated.
3. Know how the statements in the CBTI are derived and be aware of the limitations such methods may have.
4. Ensure the accuracy of test data that are hand-entered into the computer.

1. When interpreting the CBTI results, be aware of potential limitations, general and specific, to the reports being used . For example:
• Score interpretations are based on administration in proctored, standardised conditions and the test has been administered under open or controlled modes and there is no evidence provided to support the validity of the report under such conditions.
• T ests are completed in an administration mode that makes it impossible to guarantee the true identity of the test-taker.
• T ests alone, however administered, may not provide a complete assessment of an individual , as other confirmatory or ancillary information is not considered .
2. Select and use the most appropriate CBTI template for the client or intended audience.
3. Ensure that the language and information given in the CBTI fit the needs of the intended stakeholder (e.g., test-taker, organisation, and client).
4. Confirm that there is a sound basis for the CBTI and that its rationale is well-documented.
5. Where possible, edit CBTI reports to include information obtained from other sources to ensure a comprehensive treatment of the test-taker’s background, behaviour, ability, aptitude, and personality.
6. Ensure appropriate, relevant, and timely feedback is provided to the test-taker and other relevant stakeholders.
7. Ensure that Internet testing presents test interpretations in a comprehensible and meaningful form.
8. Provide client test interpretations that are appropriate for the context and intended use of the test (e.g., high or low stakes testing, corporate versus individual applications).
9. Take account of ethical issues surrounding the provision of feedback using the Internet (e.g. the difficulty of knowing the effect of providing negative feedback to a test-taker, the lack of knowledge of the emotional state of the test-taker, or the difficulty of providing immediate support to a test-taker when feedback has a negative impact). Where appropriate, feedback should include directions on how to access support and other information.

1. To monitor for possible adverse impact, collect data on the number of individuals accessing the CBT/Internet test from protected/minority groups .
• For most countries such groups may be legally defined in terms of one or more of the following: ethnicity, gender, age, disability, religion, and sexual orientation.
2. Where there is evidence of possible inequality of access, offer the use of alternative methods of testing.
3. Where possible, collect data to monitor group differences in test scores.
4. Consider the appropriateness and feasibility of Internet testing if testing in locations with limited access to computer technology or the Internet.
5. If testing internationally, use the country-specific adapted versions of the test, if available.