Political Knowledge Scale

The construct is whether a citizen can retrieve correct political information. Treating don't-know as missing would drop exactly the people who lack the knowledge, inflating the apparent average. The standard convention codes don't-know as incorrect, though researchers note this can penalize cautious respondents and contributes to observed group differences, motivating item response theory adjustments.

Delli Carpini and Keeter argued that general political knowledge is the most reliable and broadly useful measure because it correlates with knowledge across domains and predicts a wide range of outcomes. Domain-specific batteries (e.g., knowledge of a particular policy) are appropriate when studying issue-specific reasoning, but for general models of competence and engagement the balanced general battery is preferred.

The classic balanced battery uses five items spanning rules of the game, policy substance, and political actors, which provides acceptable reliability at low survey cost. Longer batteries (10-20 items) improve precision and support item response theory scaling, but five well-chosen, appropriately difficult items remain the standard compromise for omnibus surveys.