Monday, 6 July 2015

Using computerised assessments in schools: Advantages and disadvantages

As a psychologist who has been involved in the development of computerised educational assessments for many years, I am an advocate of using this test format... but what are the advantages (and disadvantages) of computerised assessment in screening for special educational needs?

There are several advantages to computerised educational tests. They are labour-saving in terms of test administration and scoring, and can be used by less highly trained personnel. The delivery of test items is more standardised and controlled (i.e. precise timing and delivery of instructions) than with human delivery and the use of a computerised test eliminates scoring errors, both of which make the test more reliable. Computerised tests are more interactive and children generally prefer them, especially pupils with SEN. Furthermore, they allow for a range of input and output devices (touch screens, Braille keyboards, joystick, mouse, speech recognition etc), making them more accessible to disabled pupils. Computers can effectively record additional data, such as response times and pattern of responses. Also, the provision of feedback is immediate and, for some tests, includes automated reports for parents which can be edited if required. It is also much easier and efficient to store computerised test records than paper-based ones. Probably the major advantage of computerised tests is that they allow for adaptive testing, where the level is adapted to suit the ability of the pupil being tested. This makes the test more efficient as pupils are not spending time on items that are too easy or too difficult for them. Finally, computerised tests MAY be less affected by sex differences. In my own research, I have found that sex differences evident on conventional literacy tests, with girls outperforming boys, were not apparent on equivalent computerised tests. The reason for this is unclear – it could be that the computerised tests are more objective and less susceptible to gender bias; however it may be due to the interaction with the computer enhancing the motivation of boys more than girls.

LASS 8-11 Report

Many of the criticisms of computerised assessment apply equally, or more so, to conventional assessment (e.g. security issues in storing test data, cost, potential misuse of tests or results etc.) or are no longer relevant criticisms (e.g. pupils not being used to using computers, lack of equivalence with paper versions of the tests etc.).  However, one disadvantage of computerised tests is that they may not pick up on certain aspects of the testing process, such as any disruptions to the test session or whether the child is anxious, distracted or rushing. Certainly some of these issues may be clear from the test results; for example a child who is rushing through a test will have fast completion rate (although not all tests provide this information) but a low score. It is, therefore, important that a tester oversees the testing process.

The advantages of computerised testing appear to outweigh the disadvantages. Computerised testing, using networked computer suites, allows for the efficient screening of all pupils in a year group. The outcome of this is that children with difficulties are less likely to slip through the net and SEN is identified at a younger age. Early identification allows for early intervention and reduces the frustration experienced by a child with unidentified learning needs.

Finally, whichever test format you choose, source tests with good evidence of their technical qualities (e.g. representative, adequately sized, appropriate UK norm groups; sufficient internal consistency; good test-retest reliability; high construct validity etc) - but that may be for a future blog...

Further reading
Horne, J. (2007). Gender differences in computerised and conventional educational tests. Journal of Computer Assisted Learning, 23, 47-55.
Singleton, C., Horne, J. & Simmons, F. (2009). Computerised screening for dyslexia in adults. Journal of Research in Reading, 32, 137-152.
Singleton, C., Thomas, K. & Horne, J. (2000). Computer-based cognitive assessment and the development of reading. Journal of Research in Reading, 23, 158-180.

Academic Bio
Dr Jo Horne is a lecturer in psychology at the University of Hull. She has been involved in the development of eight computerised assessment systems, which are now being used in schools, colleges and universities around the UK and internationally. Jo is also an editor of the British Psychological Society’s Psychological Testing Centre Test Reviews.

No comments:

Post a Comment