Chartered Financial Analyst (CFA) Practice Exam Level 2

What is a key indicator of influential cases in a dataset as per Cook's Distance?

1. Values of less than 1

2. Values of exactly 1

3. Values of 1 or greater

4. Values of less than or equal to 0

The correct answer is: Values of 1 or greater

Cook's Distance is a statistic used to identify influential observations in a regression analysis. It quantifies how much the predicted values of the model would change if a particular data point were removed. The formula for Cook's Distance incorporates both the leverage of the observation and the residual of the regression model, allowing for a comprehensive assessment of the influence of each point. When examining Cook's Distance values, a key indicator is that values of 1 or greater suggest that the data point in question is potentially influential. This level indicates that the observation significantly affects the regression model's parameters and predictions. Thus, any point presenting a Cook's Distance of 1 or more can be considered of particular interest for further investigation, as it may disproportionately impact the results of the analysis. In contrast, values significantly lower than 1 imply that the observation has little to no impact on the overall regression outcome, thus falling below the threshold of being considered influential. This provides a clear demarcation where values above this threshold warrant a closer examination.