Statistical Measures and Regression Analysis Study Notes

Scenario Analysis: Orthogonal Regression Lines
- In the context of simple linear regression, several conditions can describe the relationship between two variables:
  - Perpendicular Regression Lines: When the regression line of $Y$ on $X$ and the regression line of X$ on $Y are perpendicular to each other, it indicates that the linear correlation coefficient ( $r$ ) is equal to $0$ .
  - Coefficient of Determination ( $R^2$ ): This coefficient represents the proportion of the variance in the dependent variable that is predictable from the independent variable. It is calculated as the square of the Pearson correlation coefficient ( $r^2$ ).
  - Sign of $R^2$ : The coefficient of determination ( $R^2$ ) is always non-negative, ranging from $0$ to $1$ . Therefore, a negative coefficient of determination is mathematically impossible in standard linear regression.
  - Interpretation of No Relationship: If there is no linear relationship between variables, the correlation coefficient $r = 0$ and the coefficient of determination $R^2 = 0$ .

Case Study: Salary Adjustment in Company A
- Initial State:
  - Total workers: $n$ .
  - Salary of $n-1$ workers: $1500\,\text{euros}$ .
  - Salary of $1$ worker: $1000\,\text{euros}$ .
- Scenario Change: The worker earning $1000\,\text{euros}$ receives a raise to $1200\,\text{euros}$ .
- Impact on Gini Index:
  - The Gini Index is a measure of statistical dispersion intended to represent the income inequality or wealth inequality within a nation or any other group of people.
  - A Gini Index of $0$ represents perfect equality, while an index of $1$ represents maximal inequality.
  - Because the lowest salary in the company is being increased toward the level of the other salaries (reducing the gap between the poorest and the rest), the income distribution becomes more equitable.
  - Conclusion: The Gini Index of salaries for the company will decrease (Disminuira).

Definition and Utility:
- Moments are specific quantitative measures related to the shape of a function\'s graph. In statistics, they are used to describe the characteristics of a frequency distribution.
- Types of Moments:
  - Moments about the origin ( $a_k$ ): Used to calculate measures of position. For example, the first moment about the origin ( $a_1$ ) is the arithmetic mean ( $\bar{x}$ ).
  - Central moments ( $m_k$ ): Moments calculated about the mean. These are used to determine dispersion and shape.
- Key Applications:
  - Position: Calculated via the first moment about the origin.
  - Dispersion: The second central moment ( $m_2$ ) is the variance ( $\sigma^2$ ).
  - Shape (Skewness and Kurtosis): The third central moment ( $m_3$ ) is used to determine the asymmetry (skewness), and the fourth central moment ( $m_4$ ) is used to determine the peakiness (kurtosis) of the distribution.
- Regression Analysis: Moments also facilitate the calculation of parameters in statistical regressions, such as the Least Squares method.
- Synthesizing Statement: Moments characterize a frequency distribution and allow for the simplified calculation of position, dispersion, and shape measures.

Mathematical Relationship:
- The coefficient of determination ( $R^2$ ) and the Pearson linear correlation coefficient ( $r$ ) are related by the following formula: $R^2 = (r)^2$
- Consequently, the magnitude of the correlation coefficient is the square root of the determination coefficient: $|r| = \sqrt{R^2}$
Problem Calculation:
- Given: A regression line determined from $10$ pairs of data with a coefficient of determination $R^2 = 0.25$ .
- Step 1: Calculate the absolute value of $r$ : $|r| = \sqrt{0.25} = 0.5$
- Step 2: Determine the sign:
  - The correlation coefficient $r$ can be either positive ( $+0.5$ ) or negative ( $-0.5$ ).
  - The sign of $r$ must match the sign of the slope ( $b_1$ ) of the regression line.
- Conclusion: Since the sign of the slope or the direction of the relationship is not provided in the prompt, there is not enough information to definitively calculate the specific value of the linear correlation coefficient ( $r$ ).

The Circular (Cyclic) Property:
- This property is an extension of the time-reversal test. It applies to index numbers involving three or more periods (e.g., period $0$ , period $1$ , and period $2$ ).
- Mathematically, the circular property is satisfied if: $I_{0/1} \times I_{1/2} \times I_{2/0} = 1$
- Or, alternatively: $I_{0/2} = I_{0/1} \times I_{1/2}$
- This property allows for the shifting of the base period without requiring a recalculation of the entire series from raw data.