Survey Scales Don't Explain Ideology Gaps — Perceptions Do

🔎 What Was Reexamined

Measuring how people view politicians is complicated because respondents often use survey response scales differently (differential item functioning). The classic Aldrich and McKelvey (1977) approach corrects for scale use but does so by assuming every respondent perceives political stimuli identically. This study introduces a new modeling approach that keeps the Aldrich–McKelvey framework but adds anchoring vignettes, enabling adjustment for scale use without forcing identical perceptions across respondents.

🧭 How the Model Works

The model combines two elements to separate scale use from genuine perceptual differences:

• Builds on Aldrich and McKelvey's spatial adjustment idea while relaxing its identifying assumption.
• Uses anchoring vignettes to identify and adjust for respondents' scale use.
• Estimates individuals' underlying perceptions of political actors rather than treating all perception as common.

📊 What the Data Reveal

The approach is applied to survey data on Americans' perceptions of parties, elected officials, and other political actors. Key findings include:

• Contrary to previous arguments, most variation in ideology ratings arises from differences in underlying perceptions, not from differing scale use.
• Perceptions of Republican politicians and of the Republican Party show no significant differences by respondent partisanship.
• Democratic and Republican respondents differ strongly in their perceptions of the ideology of Democratic political actors and of the Supreme Court.

⚖️ Why It Matters

This modeling approach changes how measurement problems in political perception are interpreted: adjustments for scale use alone do not eliminate meaningful partisan differences in perceived ideology. Allowing respondents to differ in their perceived locations of political actors clarifies when survey disagreements reflect true perceptual divergence versus mere response-scale behavior, with implications for studies of public opinion, representation, and measurement practice.