Correlation and Standard Error of Estimate Explained

Understand Correlation and Standard Error of Estimate

Key rules to remember

Karl Pearson correlation coefficient

r = [nΣXY − ΣXΣY] ÷ √{[nΣX² − (ΣX)²] × [nΣY² − (ΣY)²]}

Use when you have raw data. The value always lies between -1 and +1.

Covariance form of r

r = Cov(X, Y) ÷ (σx × σy), where Cov(X, Y) = ΣXY ÷ n − X̄ × Ȳ

Use when means and standard deviations are given or easy to find. Use the same divisor (n) throughout.

Coefficient of determination

r² = Explained variation ÷ Total variation = 1 − (Unexplained variation ÷ Total variation)

Express as a percentage when interpreting. It lies between 0 and 1.

Link between r and regression coefficients

r = ±√(byx × bxy); byx = r × σy ÷ σx; bxy = r × σx ÷ σy

Take the sign of the regression coefficients. Both have the same sign as r. Their product, r², cannot exceed 1.

Standard error of estimate (definition)

Se = √[Σ(Y − Ŷ)² ÷ (n − 2)]

Divisor n − 2 is used for a simple linear regression fitted from sample data. Some questions divide by n; follow the question's instruction.

Standard error of estimate (computational)

Se = √[(ΣY² − aΣY − bΣXY) ÷ (n − 2)]

Use for Y = a + bX when the sums are given and residuals are not worked out.

Se from r

Unexplained variation = Σ(Y − Ȳ)² × (1 − r²); Se = √(Unexplained variation ÷ (n − 2))

Quick route when r and the total variation in Y are given. If using σy with divisor n, then Se = σy × √(1 − r²).

Approximate forecast range

Ŷ ± 1 Se ≈ 68% range; Ŷ ± 2 Se ≈ 95% range

A rough rule that assumes residuals are roughly normal and the sample is reasonably large. Use it only when the question asks for it.

How to solve Correlation and Standard Error of Estimate questions

1. 1Read what is asked: r, r², regression line, Se, or a forecast with its reliability. Note which variable is X (the independent one) and which is Y.
2. 2List what is given: raw pairs, or summary figures such as n, ΣX, ΣY, ΣXY, ΣX², ΣY². Build a small table with X, Y, XY, X² and Y² if you have raw data.
3. 3Total the columns and write n. Check one total again, because one slip spoils every later step.
4. 4Compute r with the formula. Check that it lies between -1 and +1 and that the sign matches the pattern of the data.
5. 5Square r to get r². Interpret it in a sentence: 'x% of the variation in Y is explained by X.'
6. 6Find the regression line Y = a + bX if needed: b = [nΣXY − ΣXΣY] ÷ [nΣX² − (ΣX)²] and a = Ȳ − bX̄.
7. 7Compute Se. Either find the residuals, square and sum them, then divide by n − 2, or use the computational formula or Unexplained variation = Total variation × (1 − r²).
8. 8State a conclusion: forecast value, Se, and whether the fit is strong enough to rely on. Give units and a clear recommendation.

Quickest way: Shortcut using total variation and r²

Common mistakes in Correlation and Standard Error of Estimate

• Treating r = 0.8 as '80% explained'.

  Students mix up r and r².

  Fix: Explained variation is r², not r. If r = 0.8, r² = 0.64, so 64% is explained.

• Concluding that a high correlation proves that X causes Y.

  Strong numbers feel like proof.

  Fix: Say that the variables move together. Cause needs reasoning beyond the number. Two variables may both be driven by a third, such as inflation.

• Dividing the squared residuals by n instead of n − 2 for Se.

  Students copy the standard deviation formula.

  Fix: For a sample-fitted simple regression, use n − 2, since two parameters (a and b) are estimated. Divide by n only if the question or its formula says so.

• Giving r a positive sign when the regression coefficients are negative.

  Taking the square root of byx × bxy gives a plain positive number.

  Fix: The sign of r is the sign of byx and bxy. If both are negative, r is negative.

• Comparing Se across different units or different levels of Y without care.

  Se looks like a pure number, so students judge it on its own.

  Fix: Se is in the units of Y. Judge it against the size of Y or against the mean. A Se of ₹5 lakh is small for sales of ₹5 crore and large for sales of ₹20 lakh.

• Arithmetic slips in the sums, so r exceeds 1 or the answer is nonsense.

  Large totals and a long table under time pressure.

  Fix: Always check that -1 ≤ r ≤ 1. Reduce the data by subtracting a common figure and re-total a column if the result looks wrong.

Worked examples