Welcome to Central Tendency and Dispersion!
Welcome to Unit 3: Statistics! When you look at any collection of numbers—such as test scores, running times, or daily temperatures—two fundamental questions immediately arise:
1. Where is the middle or typical value? (This is Central Tendency)
2. How spread out are the numbers from each other? (This is Dispersion)
Don't worry if statistical formulas look intimidating at first glance. Once you break them down into simple step-by-step recipes, you will find them straightforward to calculate and easy to master.
1. Measures of Central Tendency (Finding the Centre)
A measure of central tendency gives us a single value that represents the "centre" or "average" of a data set. In CCEA GCSE Further Mathematics, you need to be confident working with the mean, median, and mode across raw data, frequency tables, and grouped continuous data.
The Mean (\(\bar{x}\))
The mean is the arithmetic average. We add all the values together and divide by the total number of items.
Formula for Raw Data:
\(\bar{x} = \frac{\sum x}{n}\)
• \(\sum x\) means "add up all the individual values of \(x\)"
• \(n\) is the total number of data values
Formula for Frequency Tables:
When data is arranged in a frequency table with values \(x\) and frequencies \(f\):
\(\bar{x} = \frac{\sum f x}{\sum f}\)
• \(\sum f x\) means multiply each value by its frequency, then add them all together
• \(\sum f\) is the total frequency (the total number of observations, \(n\))
For Grouped Continuous Data:
When data is grouped in intervals (e.g., \(10 \le x < 20\)), we do not know the exact values. We use the midpoint of each group as our estimate for \(x\), then apply \(\bar{x} = \frac{\sum f x}{\sum f}\).
The Median and Mode
• Mode: The value (or class interval) that occurs most frequently. If looking at grouped data, the interval with the highest frequency is called the modal class.
• Median: The middle value when the data is placed in numerical order. For grouped data, the median can be estimated from a cumulative frequency graph or by interpolation.
Key Takeaway: The mean uses every single piece of data, making it very sensitive to extreme values (outliers). The median is resistant to outliers because it only depends on the position of the middle item.
2. Measures of Dispersion: Range and Interquartile Range
Knowing the average is only half the story. Imagine two classes that both have a mean test score of \(50\%\). In Class A, everyone scored between \(48\%\) and \(52\%\). In Class B, scores ranged from \(0\%\) to \(100\%\). The spread (dispersion) tells us how consistent or variable the data is!
The Range
\(\text{Range} = \text{Maximum value} - \text{Minimum value}\)
While simple to compute, the range is easily distorted by a single unusually high or low value (an outlier).
Quartiles and the Interquartile Range (IQR)
Quartiles divide an ordered data set into four equal quarters:
• Lower Quartile (\(Q_1\)): The value \(25\%\) of the way through the ordered data.
• Median (\(Q_2\)): The value \(50\%\) of the way through.
• Upper Quartile (\(Q_3\)): The value \(75\%\) of the way through.
Interquartile Range (IQR):
\(\text{IQR} = Q_3 - Q_1\)
Semi-Interquartile Range:
\(\text{Semi-IQR} = \frac{Q_3 - Q_1}{2}\)
Why use the IQR? The IQR measures the spread of the middle \(50\%\) of the data, completely ignoring extreme outliers at the top and bottom ends.
Key Takeaway: A smaller range or IQR means the data is more consistent. A larger range or IQR means the data is more varied.
3. Variance and Standard Deviation
The standard deviation is the most important measure of dispersion in Further Mathematics. It calculates the average distance of each data point from the mean.
Definitions and Formulas
• Variance (\(\sigma^2\) or \(s^2\)): The average of the squared differences from the mean.
• Standard Deviation (\(\sigma\) or \(s\)): The square root of the variance: \(\sigma = \sqrt{\text{Variance}}\).
Formula for Raw Data
\(\sigma = \sqrt{\frac{\sum x^2}{n} - \bar{x}^2}\)
Memory Aid: Think "the mean of the squares minus the square of the mean", all under a square root!
Formula for Grouped Data and Frequency Tables
\(\sigma = \sqrt{\frac{\sum f x^2}{\sum f} - \left(\frac{\sum f x}{\sum f}\right)^2}\) or \(\sigma = \sqrt{\frac{\sum f x^2}{\sum f} - \bar{x}^2}\)
Step-by-Step Calculation Guide for Frequency Tables
Follow these four simple steps when given a frequency table:
Step 1: Find the midpoint \(x\) for each interval (if the data is grouped).
Step 2: Calculate a column for \(f \times x\) and find the sum \(\sum f x\).
Step 3: Calculate a column for \(f \times x^2\) (which is \(f x \times x\)) and find the sum \(\sum f x^2\).
Step 4: Substitute these totals into the formula to find the variance and standard deviation.
Worked Example
A student records the number of books read by \(10\) classmates:
• Values of \(x\): \(1, 2, 3, 4\)
• Frequencies \(f\): \(2, 4, 3, 1\)
1. Total frequency: \(\sum f = 2 + 4 + 3 + 1 = 10\)
2. \(\sum f x = (1 \times 2) + (2 \times 4) + (3 \times 3) + (4 \times 1) = 2 + 8 + 9 + 4 = 23\)
3. Mean: \(\bar{x} = \frac{23}{10} = 2.3\)
4. \(\sum f x^2 = (1^2 \times 2) + (2^2 \times 4) + (3^2 \times 3) + (4^2 \times 1) = 2 + 16 + 27 + 16 = 61\)
5. Variance: \(\sigma^2 = \frac{61}{10} - (2.3)^2 = 6.1 - 5.29 = 0.81\)
6. Standard Deviation: \(\sigma = \sqrt{0.81} = 0.9\)
Common Mistakes to Avoid
• Squaring \(f x\) instead of \(x\): In the column for \(f x^2\), remember it is \(f \times (x^2)\), NOT \((f x)^2\).
• Forgetting to square the mean: Remember to subtract \((\bar{x})^2\), not just \(\bar{x}\).
• Stopping at the variance: If a question asks for standard deviation, always remember to take the final square root!
Key Takeaway: Variance is \(\sigma^2\). Standard deviation is \(\sigma\). A low standard deviation means data points are clustered closely around the mean.
4. Linear Transformations (Coding of Data)
Coding involves changing the scale or origin of a data set using a linear formula: \(y = ax + b\), where \(a\) and \(b\) are constants.
How Transformations Affect the Mean and Measures of Spread
Imagine every person in a class gets an extra \(5\) marks on a test (\(y = x + 5\)):
• The average score increases by \(5\).
• The difference between people's scores stays exactly the same!
Now imagine everyone's score is doubled (\(y = 2x\)):
• The average score doubles.
• The gap between scores also doubles!
General Rules for \(y = ax + b\):
1. Measures of Location (Mean, Median, Mode):
Both adding/subtracting and multiplying/dividing affect the location.
• \(\bar{y} = a\bar{x} + b\)
• \(\text{Median}_y = a(\text{Median}_x) + b\)
2. Measures of Dispersion (Standard Deviation, Variance, IQR, Range):
Adding or subtracting a constant (\(+b\)) shifts all points equally and has NO effect on spread. Only multiplying or dividing (\(\times a\)) changes the spread.
• \(\sigma_y = a \times \sigma_x\) (for \(a > 0\))
• \(\text{Variance}_y = a^2 \times \text{Variance}_x\)
• \(\text{IQR}_y = a \times \text{IQR}_x\)
• \(\text{Range}_y = a \times \text{Range}_x\)
Quick Transformation Example
Suppose a data set of heights \(x\) has a mean of \(\bar{x} = 160\text{ cm}\) and a standard deviation of \(\sigma_x = 8\text{ cm}\).
If the data is coded using \(y = \frac{x - 10}{2} = 0.5x - 5\):
• Coded Mean: \(\bar{y} = 0.5(160) - 5 = 80 - 5 = 75\)
• Coded Standard Deviation: \(\sigma_y = 0.5 \times 8 = 4\) (the \(-5\) is ignored!)
Key Takeaway: Shift (\(+b\)) moves the average, but doesn't change spread. Scale (\(\times a\)) multiplies both the average and the spread.
5. Chapter Summary and Quick Revision Checklist
Before sitting your exam, make sure you can answer YES to each of these points:
• Can you calculate the mean, median, and modal class from raw, discrete, and grouped continuous data?
• Do you know the formulas for \(\text{IQR} = Q_3 - Q_1\) and \(\text{Semi-IQR} = \frac{Q_3 - Q_1}{2}\)?
• Can you set up a table with columns \(x\), \(f\), \(fx\), and \(fx^2\) to compute \(\sum f\), \(\sum fx\), and \(\sum fx^2\)?
• Can you apply the formula \(\sigma = \sqrt{\frac{\sum fx^2}{\sum f} - \bar{x}^2}\) accurately?
• Can you apply linear coding rules: \(\bar{y} = a\bar{x} + b\) and \(\sigma_y = a\sigma_x\)?