Statistics and Single Variable Data Handling

Exam Tip: When using a scientific calculator, you can often find the standard deviation for a population ($\sigma_x$) or a sample ($s_x$) in the "Var" sub-menu of the Statistics mode.

Handling Grouped Data

Sometimes, data is too vast to list individually, so it is organized into intervals (e.g., 0–4, 5–9). When working with grouped data, you cannot use the individual numbers within the group. Instead, you must use the midpoint of the interval to represent the group.

For example, in a group ranging from 0 to 4, the midpoint is calculated as: $\frac{0 + 4}{2} = 2$

To calculate the mean or standard deviation for grouped data, input these midpoints into your calculator's frequency table as the $x$-values, and use the frequencies as the $y$-values.

Regression Analysis (Bivariate Data)

While univariate data looks at one variable, regression looks at the relationship between two variables: an independent variable ($x$) and a dependent variable ($y$).

Types of Regression

Depending on the shape of the data on a scatter plot, different mathematical models are used:

1. Linear Regression ($y = A + Bx$): Used when the relationship follows a straight line.
2. Quadratic Regression: Used for U-shaped or inverted U-shaped patterns.
3. Logarithmic Regression: Used for specific curved relationships.
4. Exponential Regression: Used when data increases or decreases at an accelerating rate.
5. AB Exponential Regression: A specific form of exponential modeling.
6. Power Regression: Used for relationships following a power law.
7. Inverse Regression: Used when one variable increases as the other decreases in a specific curved manner.

The Correlation Coefficient ($r$)

The strength and direction of a linear relationship are measured by the correlation coefficient, $r$. The value of $r$ always falls between $-1$ and $1$: $-1 \leq r \leq 1$

• Strong Positive Correlation ($r \approx 1$): As $x$ increases, $y$ increases.
• Strong Negative Correlation ($r \approx -1$): As $x$ increases, $y$ decreases.
• No Relationship ($r \approx 0$): The points are scattered randomly with no discernible pattern.
• Non-linear Correlation: The points follow a curve rather than a straight line.

Making Projections: Interpolation vs. Extrapolation

Once you have determined the line of best fit ($y = A + Bx$), you can use it to predict values.

• Interpolation: Predicting a value within the range of your original data points. This is generally considered reliable.
• Extrapolation: Predicting a value outside the range of your original data. This is riskier because you are assuming the trend continues indefinitely.

Calculator Pro-Tip: To find a predicted $y$ value for a given $x$ on a Casio calculator, use the "Reg" menu and select the option for "Estimated value of $y$" (often option 5).

Summary of Calculator Functions

For students using a scientific calculator (like the Casio fx-991ZA PLUS II), the STAT mode is an essential tool for efficiency.