Practice Effect & Carry Over Effect Definition & Examples

Design of Experiments >

practice effect
Practice effect can occur when participants are tested more than once. Image: Cpl. Khalil Ross|Wikimedia Commons

Practice Effect Definition

Practice effects are influences on test results when a test is taken more than once. As a simple example, a practice effects occurs when you take multiple practice SAT exams; practice can increase your overall score. However, in psychological and educational testing, the practice effect is usually an unwanted influence.

For example, unwanted practice effects are thought to affect the Wechsler Intelligence Scale for Children — Fourth Edition (WISC-IV) if the test is readministered within two years. If the test is readministered within this time, the child’s score may be inflated, giving invalid results.

Practice effects in repeated measures design

In repeated measures design, each participant is measured for multiple conditions in an experiment. For example, a group of people might be given extra help to see if it improves their math ability, and then they might be given access to an online help program. A math test is given to the participants after each treatment. Several factors contribute to the possibility of practice effect in this case, including:

  • Participants are aware of the questions on the test, so may become more adept simply because they are repeating the test.
  • Participants may become bored or tired of test taking.

Carryover effects

A carryover effect is a type of practice effect that occurs because the results from one test influences another. Carryover effects typically fall into three categories:

  • Participants receive instructions in one level that can help (or hinder) them in another level. For example, let’s say you were conducting a test to see if reading through an entire exam, before putting pen-to-paper, can increase test scores. Participants are given a second test, with no instruction to read thoroughly. They may read thoroughly anyway because of the previous instruction.
  • Participants can become aware of the purpose of the experiment during the first treatment, and this may influence their behavior in subsequent treatments.
  • One level of the independent variable can have an explicit effect on other levels. For example, a group of students is tested to see if caffeine helps with standardized test scores. If the group is given caffeine before the first test, that caffeine may affect their performance on subsequent tests.

Comments? Need to post a correction? Please Contact Us.