Selecting response anchors with equal intervals for summated rating scales

Area of Science:

• Organizational Research
• Psychometrics
• Survey Methodology

Background:

• Summated rating scales are widely used in organizational research, yet guidelines for selecting response anchors are less developed.
• Ensuring equal-interval properties in survey data is vital for parametric statistical analyses.
• Previous research has not sufficiently addressed the conceptual distance between response anchors.

Purpose of the Study:

• To identify common contexts and anchors used in summated rating scales.
• To provide empirical data on the conceptual distances between response anchors.
• To examine the impact of anchor properties on measurement and statistical outcomes.

Main Methods:

• Analysis of common contexts and anchors in summated rating scales.
• Empirical estimation of conceptual distances between response anchors.
• Experimental studies on the effects of equal-interval versus unequal-interval anchors on correlations.

Main Results:

• Common contexts include agreement, similarity, frequency, amount, and judgment.
• Provides means, standard deviations, and distribution overlaps for common anchors.
• Correlations are consistently weaker when using unequal-interval anchors compared to equal-interval anchors.

Conclusions:

• Empirical data on anchor properties can guide researchers in selecting scales that minimize measurement error.
• Using equal-interval anchors enhances confidence in statistical analyses and research findings.
• The choice of response anchors significantly impacts the strength of observed relationships between constructs.