01 · Explore
Introduction to Correlation
Correlation analysis is a statistical tool used to examine the systematic relationship between two variables.
In previous studies, we focused on summary measures for single variables. Correlation moves beyond this by asking if a change in one variable is associated with a change in another. For example, as summer temperatures rise, ice-cream sales typically increase. Conversely, as the supply of a vegetable like tomatoes increases in a mandi, its price often drops.
It is vital to distinguish between correlation and causation. Correlation measures 'covariation'—how variables move together—but it does not prove that one causes the other. Some relationships are mere coincidences, such as the arrival of migratory birds and local birth rates. Others may be driven by a third 'hidden' variable, such as rising temperatures causing both higher ice-cream sales and higher swimming-related accidents.
Pause & Try
Think it through first. Writing and checking your answer is free with an account.
Question
Does a high correlation between two variables imply that one causes the other?
Sign in to see the answer
Write your own answer and compare it with ours. It’s free.
Sign inNew here? Sign up freeNCERT reference: chapter PDF pages 1, 2.
02 · Explore
Types of Correlation
Correlation is classified based on the direction and nature of the movement between variables.
Positive Correlation occurs when both variables move in the same direction. For instance, as household income rises, consumption generally rises. If income falls, consumption falls. Both variables increase or decrease together.
Negative Correlation occurs when variables move in opposite directions. A classic economic example is the relationship between price and demand: as the price of apples rises, the quantity demanded typically decreases. Similarly, spending more time studying might correlate with a decrease in the chances of failing.
Linear vs. Non-linear: A relationship is linear if the relative movement of variables can be represented by a straight line. If the relationship follows a curve, it is non-linear. Karl Pearson's coefficient is specifically designed for linear relationships.
Pause & Try
Think it through first. Writing and checking your answer is free with an account.
Question
Give an example of a negative correlation from daily life.
Sign in to see the answer
Write your own answer and compare it with ours. It’s free.
Sign inNew here? Sign up freeNCERT reference: chapter PDF pages 2, 3, 4.
03 · Explore
Visualizing Data: Scatter Diagrams
A scatter diagram is a visual technique for examining the relationship between two variables without calculating numerical values.
In a scatter diagram, values of two variables are plotted as points on a graph. The overall direction and closeness of these points indicate the nature of the relationship. If the points cluster around an upward-sloping line, the correlation is positive. If they cluster around a downward-sloping line, it is negative.
The degree of correlation is shown by the dispersion of points. If all points lie exactly on a straight line, the correlation is 'perfect' (unity). If the points are widely scattered with no discernible slope, there is no correlation. Scatter diagrams are useful because they can reveal both linear and non-linear patterns that numerical formulas might miss.
| Pattern | Interpretation |
|---|---|
| Points rising from left to right | Positive Correlation |
| Points falling from left to right | Negative Correlation |
| Points form a perfect straight line | Perfect Correlation (r = 1 or -1) |
| Points are randomly scattered | No Correlation (r = 0) |
Pause & Try
Think it through first. Writing and checking your answer is free with an account.
Question
What does it mean if all points on a scatter diagram lie exactly on a downward-sloping straight line?
Sign in to see the answer
Write your own answer and compare it with ours. It’s free.
Sign inNew here? Sign up freeNCERT reference: chapter PDF pages 3, 4, 5.
04 · Explore
Karl Pearson’s Coefficient of Correlation
Also known as the product-moment correlation coefficient, this measure provides a precise numerical value for linear relationships.
The coefficient, denoted as 'r', measures the direction and intensity of the linear association between two variables, X and Y. It is calculated using the deviations of items from their respective arithmetic means. The sign of the correlation is determined by the covariance of the two variables.
The formula for r using raw data is: r = [NΣXY - (ΣX)(ΣY)] / √[NΣX² - (ΣX)²] * √[NΣY² - (ΣY)²]. This is identified as formula (4) in the source text and is used to calculate the coefficient directly from the observed values without first finding deviations.
Pause & Try
Think it through first. Writing and checking your answer is free with an account.
Question
What is the mathematical range of the correlation coefficient 'r'?
Sign in to see the answer
Write your own answer and compare it with ours. It’s free.
Sign inNew here? Sign up freeNCERT reference: chapter PDF pages 4, 6.
05 · Explore
Properties of the Correlation Coefficient
Understanding the properties of 'r' helps in the correct interpretation of statistical data.
First, 'r' is a pure number and has no units. Whether you measure height in feet or meters, the correlation coefficient between height and weight remains the same. Second, a value of r = 0 does not necessarily mean the variables are independent; it only means there is no 'linear' relationship. A strong non-linear relationship might still exist.
A crucial property is that 'r' is independent of the change of origin and change of scale. This means if you subtract a constant from all values (change of origin) or divide them by a constant (change of scale), the value of 'r' remains unchanged. This property is exploited in the 'step deviation method' to simplify calculations involving large numbers.
Pause & Try
Think it through first. Writing and checking your answer is free with an account.
Question
If you change the unit of measurement for X from kilograms to grams, how will the value of r change?
Sign in to see the answer
Write your own answer and compare it with ours. It’s free.
Sign inNew here? Sign up freeNCERT reference: chapter PDF pages 6, 7.
06 · Explore
The Step Deviation Method
When data values are large, the step deviation method simplifies the calculation of the correlation coefficient.
This method involves transforming the variables X and Y into smaller values, U and V, using assumed means (A and C) and common factors (h and k). The transformation is: U = (X - A) / h and V = (Y - C) / k.
Because 'r' is independent of origin and scale, the correlation coefficient between the original variables (rXY) is exactly equal to the correlation coefficient between the transformed variables (rUV). This allows us to work with much smaller numbers, reducing the risk of arithmetic errors.
Variable Transformation
U = (X - 100) / 10
If X values are 120, 150, and 190, we can use an assumed mean of 100 and a common factor of 10. The transformed U values become (120-100)/10 = 2, (150-100)/10 = 5, and (190-100)/10 = 9. These smaller values are easier to square and multiply.
Pause & Try
Think it through first. Writing and checking your answer is free with an account.
Question
Why is the step deviation method used in statistics?
Sign in to see the answer
Write your own answer and compare it with ours. It’s free.
Sign inNew here? Sign up freeNCERT reference: chapter PDF pages 9, 10.
07 · Explore
Spearman’s Rank Correlation
Spearman's rank correlation is used when variables cannot be measured quantitatively but can be ranked in order of preference or quality.
Developed by C.E. Spearman, this method is ideal for qualitative attributes like beauty, honesty, or intelligence. It is also useful when data contains extreme values (outliers) that might distort Karl Pearson's coefficient, or when the relationship is non-linear but monotonic (moving in one direction).
The formula for Spearman's rank correlation (rs) is: rs = 1 - [6 * ΣD² / (n³ - n)], where 'D' is the difference between the ranks assigned to the two variables for each observation, and 'n' is the number of observations.
Calculating Rank Correlation
rs = 1 - [6 * ΣD² / (n³ - n)]
For 5 competitors (n=5) with a sum of squared rank differences (ΣD²) of 14: rs = 1 - [6 * 14 / (5³ - 5)] = 1 - [84 / 120] = 1 - 0.7 = 0.3. This shows a weak positive relationship between the rankings.
Pause & Try
Think it through first. Writing and checking your answer is free with an account.
Question
When is Spearman's rank correlation preferred over Karl Pearson's coefficient?
Sign in to see the answer
Write your own answer and compare it with ours. It’s free.
Sign inNew here? Sign up freeNCERT reference: chapter PDF pages 10, 11, 12.
08 · Explore
Handling Repeated Ranks
When two or more items have the same value, they are assigned a common rank, requiring a correction in the Spearman formula.
If multiple observations have the same value, they are assigned the average of the ranks they would have occupied. For example, if the value 50 appears at the 9th, 10th, and 11th positions in a sorted list, all three are assigned the average rank of 10 [(9+10+11)/3].
To account for these ties, a correction factor [(m³ - m) / 12] is added to ΣD² for each group of tied ranks, where 'm' is the number of items in that tie. The modified Spearman formula is: rs = 1 - [6 * (ΣD² + Σ(m³ - m) / 12) / (n(n² - 1))]. This ensures the correlation remains accurate despite the lack of unique rankings.
Correction Factor Calculation
Correction = (m³ - m) / 12
If the value '50' appears 3 times in a dataset, m = 3. The correction factor is (3³ - 3) / 12 = (27 - 3) / 12 = 24 / 12 = 2. If another value appears twice, m = 2, and its correction is (2³ - 2) / 12 = 0.5. The total correction added to ΣD² would be 2.5.
Pause & Try
Think it through first. Writing and checking your answer is free with an account.
Question
How do you assign ranks if three students score the same marks and occupy the 9th, 10th, and 11th positions?
Sign in to see the answer
Write your own answer and compare it with ours. It’s free.
Sign inNew here? Sign up freeNCERT reference: chapter PDF pages 13, 14.