Assessing Multi-Item Scales for Subjective Measurement