Reliability

• Reliability refers to the consistency and precision of a test.
• Errors are factors other then what the test aims to measure that may influence performance or results.
• The Reliability Coefficient is the ratio between true score variance and total variance. It is the index of a test's reliability.
• TYPES OF ERROR

  • Measurement Error
  • Random Error
  • Systematic Error
• TYPES OF RELIABILITY

  • Test-Retest Reliability
  • Parallel or Alternate Forms Reliability
  • Internal Consistency
  • Split-half Reliability
  • Interscorer Reliability
• Test-Retest Reliability is an estimate of reliability obtained by correlating pairs of scores from the same people on two different administrations of the test.
• Carryover effect happens when the test-retest interval is short, which results to the second test being influenced by what they practiced or remembered from the first test.
• Practice effect is when the test taker's score on the second test is higher than the first test due to their experience on the first test.
• Parallel or Alternate Form Reliability is established when when at least two versions of the same test yield almost the same results.
• Split-half Reliability is obtained when correlating two pairs of scores obtained from equivalent halves of the same test administered once.
• Internal Consistency or Inter-item Reliability is used when items are administered once and is the degree to which each item measures the same construct.
• Inter-scorer Reliability refers to the degree of agreement or consistency between two or more scorers with regards to a particular measure.
• Difficulty is the quality of a test item of being not easily solved or accomplished.
• Discrimination is the degree in which a test item can differentiate among people with higher or lower levels of the trait.