Methods of Constructing Indexes

An index is a composite measure that summarizes and ranks several indicators (variables/items) to represent a complex construct (e.g., knowledge, socio-economic status, adoption, modernization).

Since one variable alone cannot capture the whole concept, index construction combines multiple variables into a single score.

1. Arbitrary or Equal Weight Method

• Researcher selects relevant items based on judgment/experience.
• All items are assigned equal weights.
• Simple but subjective.

Formula:

Index = Number of practices adopted / Total practices recommended × 100

• Example: Farmer Adoption Index = (Practices adopted ÷ Total practices recommended) × 100.
• Most commonly used in Extension for Adoption/Knowledge Index.

2. Weighted Index Method

• Items are assigned different weights according to their importance.
• Weights may be decided by experts or by statistical methods (e.g., factor analysis, regression).

Formula: Index = Σ(Wi×Xi) / ΣWi

Where: WiW_iWi​ = weight of item and XiX_iXi​ = score of respondent on item

• Example: Socio-Economic Status (SES) Index = (Education × 0.4) + (Occupation × 0.3) + (Income × 0.3).
• More scientific than equal weight; removes bias by considering item importance.

3. Thurstone’s Equal Appearing Interval Method (1928)

• Panel of judges rates items on a continuum (favorable → unfavorable).
• Median values are used as item scores.
• Items are equally spaced along the continuum.

• Use: Mostly for attitude scaling, but principle applies in index construction.
• Thurstone = Equal Appearing Interval Scale.

4. Likert’s Summated Rating Method (1932)

• Respondents express agreement/disagreement on a 5- or 7-point scale.
• Scores are summed across items → total score = index value.

• Example: Attitude Index toward Organic Farming.
• Most popular scaling method in social sciences.

5. Cumulative Scaling (Guttman Scalogram Analysis)

• Items are arranged in hierarchical order of difficulty/intensity.
• Endorsement of a stronger item implies endorsement of weaker items.
• Produces a cumulative index.
•  
• Example: Technology adoption index – if a farmer uses advanced hybrid seed, it is assumed he uses basic improved seed also.
• Check for scalability using coefficient of reproducibility (>0.90 acceptable).

6. Z-Score or Standard Score Method

• Raw scores are standardized to z-scores to eliminate unit differences.
• Composite index = sum/average of z-scores.
• Formula: Z = X−XˉSD
• ASRB Point: Used when variables are in different units (e.g., income in ₹, land in acres, education in years).

7. Factor Analysis / Principal Component Analysis (PCA)

• Advanced multivariate statistical method.
• Reduces large number of variables into a smaller number of factors/components.
• Weights (factor loadings) are used to construct an index.

• Example: Human Development Index (HDI), Composite Development Indices.
• Statistically strongest method; widely used in modern extension & development studies.

8. Adoption & Knowledge Indexes (Special Cases in Extension)

• Knowledge Index = (Correct responses ÷ Total possible responses) × 100.
• Adoption Index = (Practices adopted ÷ Practices recommended) × 100.
• Extent of Adoption = Σ (Adopted score ÷ Maximum possible score) × 100.

ASRB Point: Frequently asked in Extension Education exams.

Summary Table for Quick Revision