Chapter-6

Chapter 6: Test of Validity, Reliability and Normality (8 hours)

  • Importance of testing for reliability, validity, and normality in social science research:

    • Ensures credibility and robustness of findings.

    • Acquires high-quality, credible results.

    • Determines accuracy, consistency, and appropriateness of measurements.

Lesson 1: Validity and Reliability (4 hours)

Objectives:

  1. Define validity and reliability of tests.

  2. Discuss different approaches to validity.

  3. Present methods for solving reliability of tests.

  4. Compute validity and reliability coefficients.

  5. Interpret validity and reliability coefficients.

Validity:

  • Crucial for measuring intended constructs in research.

  • Without valid measures, conclusions may be misleading.

  • Types of validity:

    1. Content Validity:

      • Evaluates how well a measure captures the theoretical idea.

      • Ensured through expert feedback (SMEs).

      • Content Validity Ratio (CVR) is calculated to assess validity.

    2. Criterion Validity (Criterion-related validity):

      • Measures how well a test predicts or correlates with the construct it measures.

      • Types:

        • Concurrent Validity: Assessed at the same time.

        • Predictive Validity: Assessed at a future time.

Lesson 2: Test of Normality (4 hours)

Objectives:

  1. Define normality and its relevance in statistical analysis.

  2. Use graphical methods to assess normality:

    • Histograms

    • Q-Q plots

  3. Use numerical methods to assess normality:

    • Skewness and Kurtosis

    • Shapiro-Wilk test

Normality Test:

  • Determines if data exhibit a normal distribution.

  • Important for appropriate use of statistical tools (correlation, regression, etc.).

  • Methods for assessing normality:

    1. Graphical:

      • Histograms: Visual representation.

      • Q-Q Plots: Compare probability distributions.

    2. Numerical:

      • Skewness: Measures asymmetry.

      • Kurtosis: Measures the height and sharpness of the distribution.

      • Shapiro-Wilk Test: Tests the null hypothesis of normality.

Reliability:

  • Ensures consistency and stability of measurements over time.

  • Types of reliability methods:

    1. Inter-Rater Reliability: Consistency of scores among multiple raters.

    2. Internal Consistency Reliability: Degree of consistency across items measuring the same construct (e.g., Cronbach's Alpha, KR-20).

Practice and Example:

  • Applying CVR, CVI, and reliability testing in sample instruments.

  • Statistical analysis methods and results interpretation.