The difference between the observed value and the model's predicted value for a single observation is called:

The difference between the observed value and the model's predicted value for a single observation is called:

• A. Residual
• B. Residuals
• C. Reliability
• D. Robust test

Answer: (A)

In regression, the difference between what you actually observe for a single case and what the model predicts for that same case is called a residual. It’s the error for that observation, computed as observed minus predicted. When you look at all observations, you refer to the set of differences as residuals, but for one data point the term is residual. Small residuals mean the model predicts that case well; large residuals suggest the model misses that case or indicates potential outliers and can flag issues with assumptions like constant variance. Reliability and robust tests describe other concepts (measurement consistency and methods that tolerate assumption violations, respectively) and don’t describe the single-case discrepancy in a regression model.

In regression, the difference between what you actually observe for a single case and what the model predicts for that same case is called a residual. It’s the error for that observation, computed as observed minus predicted. When you look at all observations, you refer to the set of differences as residuals, but for one data point the term is residual. Small residuals mean the model predicts that case well; large residuals suggest the model misses that case or indicates potential outliers and can flag issues with assumptions like constant variance. Reliability and robust tests describe other concepts (measurement consistency and methods that tolerate assumption violations, respectively) and don’t describe the single-case discrepancy in a regression model.