How does adding a constant $k$ to every value in a dataset affect the mean and the standard deviation?

The mean increases by the constant $k$, as the entire distribution shifts. The standard deviation remains unchanged because the relative distances between data points do not change.

When data is coded using $y = \frac{x - 100}{5}$, how do you retrieve the original mean $\bar{x}$ from the coded mean $\bar{y}$?

You must reverse the operations by multiplying the coded mean by 5 and then adding 100. The formula is $\bar{x} = 5\bar{y} + 100$.

If a dataset is multiplied by $-3$, what happens to the standard deviation?

The standard deviation is multiplied by $|-3|$, which is $3$. Since standard deviation measures distance/spread, it must always be a non-negative value.

What is the difference between the effect of coding on variance versus standard deviation?

Standard deviation is multiplied by the modulus of the scale factor $|a|$, whereas variance is multiplied by the square of the scale factor $a^2$.

What is an 'assumed mean' in the context of coding?

An assumed mean is a constant value subtracted from all data points to simplify calculations. It effectively shifts the data set so that the values are smaller and easier to sum or square.

Why is the variance of $(x - a)$ equal to the variance of $x$?

Subtracting a constant $a$ is a translation that shifts every data point by the same amount. Because variance measures the spread of data relative to the mean, and that spread remains constant during a shift, the variance is unchanged.

What common mistake occurs when reversing the coding $y = 2x + 10$ to find the original mean?

Students often divide by 2 before subtracting 10. The correct algebraic reversal of $y = 2x + 10$ requires subtracting 10 first, then dividing the result by 2.

Define the linear coding formula and its components.

The formula is $y = ax + b$, where $x$ is the original value, $y$ is the coded value, $a$ is the scaling factor (multiplier), and $b$ is the translation constant (added/subtracted).

If the range of $x$ is 20 and $y = 3x - 5$, what is the range of $y$?

The range of $y$ is $3 \times 20 = 60$. Like standard deviation, range is a measure of spread and is only affected by the multiplier, not the subtraction of 5.

How does dividing every value in a dataset by 10 affect the variance?

The variance is divided by $10^2$, which is 100. This is because variance is a squared measure of spread, so any scaling factor must be squared.

Library Podcasts

Courses

Referral & Rewards

Revision Notes

AS-Level

Cambridge International Examinations

Maths

Probability And Statistics 1

Data Presentation & Interpretation

Coding Data

Summary

Coding is a statistical technique used to simplify complex datasets by applying linear transformations. By adding, subtracting, multiplying, or dividing all values in a set by a constant, researchers can work with more manageable numbers without losing the underlying statistical relationships. Understanding how these transformations affect measures of location (like the mean) versus measures of spread (like standard deviation) is fundamental to accurate data analysis.

1. Definition & Core Concepts

Coding is the process of transforming a dataset using a mathematical formula to make calculations easier or to standardize data for comparison.

The most common form of coding is a linear transformation, expressed by the formula $y = ax + b$ , where $x$ represents the original data and $y$ represents the coded data.

An assumed mean is a specific type of coding where a constant value is subtracted from every data point to center the data around zero or a smaller number, simplifying the calculation of the actual mean.

Diagram showing a distribution curve shifting along the x-axis due to additive coding, illustrating that the shape and spread remain identical while the location changes.

2. Impact on Measures of Location

3. Impact on Measures of Spread

4. Key Distinctions

5. Summary Statistics with Assumed Mean

6. Exam Strategy & Tips

Coding Data

Summary

1. Definition & Core Concepts

Coding is the process of transforming a dataset using a mathematical formula to make calculations easier or to standardize data for comparison.

The most common form of coding is a linear transformation, expressed by the formula $y = ax + b$ , where $x$ represents the original data and $y$ represents the coded data.

Diagram showing a distribution curve shifting along the x-axis due to additive coding, illustrating that the shape and spread remain identical while the location changes.

2. Impact on Measures of Location

Measures of location, such as the mean, median, and mode, are directly affected by every part of the coding formula.

If the data is coded as $y = ax + b$ , the new mean $\bar{y}$ is calculated by applying the same operations to the original mean: $\bar{y} = a\bar{x} + b$ .

To retrieve the original mean from coded results, you must reverse the operations in the opposite order: $\bar{x} = \frac{\bar{y} - b}{a}$ .

3. Impact on Measures of Spread

Measures of spread, such as standard deviation and range, describe the distance between data points and are only affected by multiplication or division.

Adding or subtracting a constant ( $b$ ) does not change the spread because every point moves by the same amount, keeping the distances between them identical.

When multiplying by a constant ( $a$ ), the standard deviation scales by the modulus (absolute value) of that constant: $\sigma_y = |a|\sigma_x$ .

The variance is affected by the square of the multiplier, meaning $\sigma_y^2 = a^2 \sigma_x^2$ .

4. Key Distinctions

It is vital to distinguish between how location and spread respond to additive versus multiplicative changes.

Statistic	Addition/Subtraction ( $+b$ )	Multiplication/Division ( $\times a$ )
Mean / Median	Changes by $+b$	Changes by $\times a$
Std. Deviation	No Change	Changes by $
Variance	No Change	Changes by $a^2$

5. Summary Statistics with Assumed Mean

When using an assumed mean $a$ , the coded data is often represented as $(x - a)$ . The sum of this coded data is $\sum(x - a)$ .

The variance of the original data can be calculated directly from the coded summary statistics: $Var(x) = \frac{\sum(x-a)^2}{n} - (\frac{\sum(x-a)}{n})^2$ .

Note that because subtraction does not affect spread, the variance of $(x-a)$ is exactly the same as the variance of $x$ .

6. Exam Strategy & Tips

Check the Multiplier: Always use the absolute value of the multiplier when adjusting the standard deviation; spread can never be negative.
Reverse Operations: When finding the original mean, ensure you solve the equation $\bar{y} = a\bar{x} + b$ for $\bar{x}$ correctly by subtracting $b$ before dividing by $a$ .
Units Matter: Remember that standard deviation maintains the original units, while variance uses squared units. Coding must respect these dimensions.
Sanity Check: If you add 10 to every score in a test, the average should go up by 10, but the 'gap' between the highest and lowest student should remain the same.