Q.70 The correlation between the observed and predicted values of the outcome variable is known as –
1. Coefficient of determination (R2)
2. Multiple correlation coefficient
3. Cox regression
4. Multiple linear regression
R² Measures Observed vs Predicted Fit: Q.70 Correct Answer Identified
In statistical modeling, the correlation between observed and predicted outcome values quantifies model accuracy. This metric specifically gauges how well predictions match actual data points.
Correct Answer
Option 1: Coefficient of determination (R²) directly represents this correlation as the squared Pearson correlation (r) between observed and predicted values.
Option Explanations
-
Option 1 correct: R² = 1 – (SS_res / SS_tot) equals cor(observed, predicted)², measuring shared variance between actual outcomes and model predictions.
-
Option 2 incorrect: Multiple correlation coefficient (R) measures linear association between one dependent variable and multiple predictors, not specifically observed vs predicted.
-
Option 3 incorrect: Cox regression is a survival analysis model for time-to-event data using partial likelihood, not a correlation measure.
-
Option 4 incorrect: Multiple linear regression is the modeling technique producing predictions, not the correlation metric itself.
Key Statistical Insight
R² ranges 0-1; higher values indicate better fit. In simple linear regression, R² perfectly equals the squared simple correlation coefficient.
Exam Strategy
Link “correlation observed-predicted” directly to R² definition. Distinguish R (correlation) from R² (variance explained) for bioinformatics/stats questions.
