Reliability and Validity of Research Instruments

1. Reliability

Meaning

• Reliability refers to the consistency, stability, and dependability of a measuring instrument.
• A reliable instrument will give the same results when applied repeatedly under similar conditions.

Kerlinger (1986): “Reliability is the accuracy or precision of a measuring instrument.”

Types of Reliability

• Test–Retest Reliability; Same test administered to the same group after some time. Measures stability over time. Example: Administering an adoption scale to farmers twice in a month.
• Split-Half Reliability; Test divided into two halves (odd-even, first–second half). Correlation between halves indicates internal consistency.
• Parallel-Form (Equivalent Form) Reliability; Two equivalent versions of the test administered to the same group. Example: Two versions of a knowledge test on crop practices.
• Inter-Rater / Inter-Observer Reliability; Agreement between different observers/raters. Example: Two extension agents rating farmer participation in training.

Key Point

• Reliability = Consistency of measurement
• Symbolically measured by correlation coefficient (r), usually reliability ≥ 0.70 is acceptable.

2) Validity

Meaning

• Validity is the extent to which an instrument measures what it is supposed to measure.
• A valid instrument ensures accuracy and truthfulness of results.

Kerlinger (1986): “Validity is the extent to which an instrument measures what it claims to measure.”

Types of Validity

• Content Validity; Degree to which test items represent the entire domain of the construct. Example: Knowledge test on Integrated Pest Management (IPM) should cover all aspects (insect pests, methods, chemicals, biological control).
• Construct Validity; Degree to which the instrument actually measures the theoretical construct. Checked through factor analysis or correlation with related variables. Example: An attitude scale should truly reflect attitude, not just knowledge.
• Criterion-Related Validity; Degree to which instrument correlates with an external criterion.
  • Two forms:
    • Concurrent validity: Instrument correlates with an existing standard at the same time.
    • Predictive validity: Instrument predicts future performance (e.g., entrance exam predicting academic success).
• Face Validity (Weakest form); The instrument appears to measure what it should, judged by experts.
  • More subjective than statistical.

Key Point

• Validity = Accuracy of measurement
• An instrument can be reliable but not valid (e.g., consistently giving wrong results).
• But if it is valid, it must also be reliable.

Relationship Between Reliability and Validity

• Reliable but not valid: A weighing machine consistently shows +5 kg error.
• Valid but not reliable: Impossible, because inconsistency cannot lead to true measurement.
• Best instrument: Both reliable and valid.

Exam-Ready Summary

One-liner for exams: Reliability is the consistency of a research instrument, while validity is its accuracy. Reliability is necessary but not sufficient for validity.