In density forecasts, respondents are asked to assign probabilities to pre-specified ranges of inflation. We show in two large-scale experiments how responses vary when we modify the response scale. Asking an identical question with modified response scales induces different answers: Shifting, compressing or expanding the scale leads to shifted, compressed and expanded forecasts. Mean forecast, uncertainty, and disagreement can change by several percentage points. We discuss implications for survey design and how central banks can adjust the response scales during times of high inflation.