Multiple Linear Regression

Econometrics

Contents

1 Introduction: From Simple to Multiple Regression 2

2 The Classical Assumptions 2

3 Estimation by Ordinary Least Squares (OLS) 2

3.1 Matrix Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3.2 Deviation-Form Normal Equations (Two Explanatory Variables) . . . . . . . . . . . 3

4 Goodness of Fit 3
4.1 Decomposition of Variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
4.2 Coefficient of Multiple Determination, R2 . . . . . . . . . . . . . . . . . . . . . . . . 3
4.3 Adjusted R2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

5 Testing the Overall Significance of the Model: The F-Test 4

6 ANOVA (Analysis of Variance) Table 4

7 Variance of the Error Term 4

8 Testing Individual Coefficients: The t-Test 4

9 Multicollinearity 5

10 Example 1 6

11 Example 1 — Matrix Approach 7

12 Example 2 8

13 Example 3: Detecting Multicollinearity 9

14 Summary of Key Formulas 9

15 Past Paper Style Questions 11

Multiple Linear Regression Lecture Notes

1 Introduction: From Simple to Multiple Regression

Simple linear regression explains Y using a single explanatory variable X. In reality, most outcomes
are influenced by several factors simultaneously. Multiple linear regression extends the model
to two or more explanatory variables:

Yi = β0 + β1 X1i + β2 X2i + · · · + βk Xki + εi

where:

• Yi = dependent (response) variable for observation i

• X1i , X2i , . . . , Xki = the k explanatory (independent) variables
• β0 = intercept; β1 , . . . , βk = partial slope coefficients
• εi = random error term

Key Idea: Partial (Ceteris Paribus) Interpretation

Each βj measures the change in Y for a one-unit increase in Xj , holding all other explanatory
variables constant. This ”netting out” of the effects of other regressors is the central advantage
of multiple regression over running several separate simple regressions.

Examples of multiple regression models:

• Sales = f (advertising expenditure, number of salespeople)

• House price = f (size, number of bedrooms, distance to city centre)

• Wage = f (years of education, years of experience, gender)

2 The Classical Assumptions

1. Linearity: The model is linear in the parameters (β0 , . . . , βk ).

2. Zero mean error: E(εi ) = 0.
3. Homoscedasticity: V ar(εi ) = σ 2 for all i (constant error variance).
4. No autocorrelation: Cov(εi , εj ) = 0 for i ̸= j.
5. No perfect multicollinearity: No exact linear relationship exists among the explana-
tory variables.
6. Exogeneity: X’s are uncorrelated with εi .
7. Normality (for inference): εi ∼ N (0, σ 2 ).
8. n > k + 1: the number of observations must exceed the number of parameters estimated.

3 Estimation by Ordinary Least Squares (OLS)

3.1 Matrix Formulation

For k explanatory variables and n observations, write the model as:

Y = Xβ + ε

2
Multiple Linear Regression Lecture Notes

where Y is n × 1, X is n × (k + 1) (including a column of 1’s for the intercept), β is (k + 1) × 1,

and ε is n × 1.

OLS Estimator (Matrix Form)

β̂ = (X′ X)−1 X′ Y

This minimises SSE = ε′ ε = e2i , the sum of squared residuals.

P

3.2 Deviation-Form Normal Equations (Two Explanatory Variables)

For the common exam case of two explanatory variables (k = 2), it is far easier to work with
deviations from the mean rather than full matrix inversion. Let x1 = X1 − X̄1 , x2 = X2 − X̄2 ,
y = Y − Ȳ . The slope coefficients solve:
X X X
b1 x21 + b2x1 x2 = x1 y
X X X
b1 x1 x2 + b 2 x22 = x2 y
Solve simultaneously (e.g. by Cramer’s Rule):

( x1 y)( x22 ) − ( x2 y)( x1 x2 ) ( x21 )( x2 y) − ( x1 x2 )( x1 y)

P P P P P P P P
b1 = b2 =
( x21 )( x22 ) − ( x1 x2 )2 ( x21 )( x22 ) − ( x1 x2 )2
P P P P P P

b0 = Ȳ − b1 X̄1 − b2 X̄2

This is exactly what the matrix formula β̂ = (X′ X)−1 X′ Y produces for k = 2 — it is simply
written out algebraically instead of via matrix inversion, which is much faster by hand.

4 Goodness of Fit

4.1 Decomposition of Variation

X X X
(Yi − Ȳ )2 = (Ŷi − Ȳ )2 + (Yi − Ŷi )2
| {z } | {z } | {z }
SST SSR SSE

(Total Sum of Squares = Regression/Explained Sum of Squares + Error/Residual Sum of Squares)

4.2 Coefficient of Multiple Determination, R2

SSR SSE
R2 = =1− 0 ≤ R2 ≤ 1
SST SST
R2 is the proportion of total variation in Y jointly explained by all the explanatory variables.

4.3 Adjusted R2

Because R2 mechanically rises whenever any variable is added — useful or not — the adjusted
R2 penalises for the number of regressors:
 
2 2 n−1
R̄ = 1 − (1 − R )
n−k−1

3
Multiple Linear Regression Lecture Notes

where k = number of explanatory variables (not counting the intercept).

R̄2 is the correct criterion for comparing models with different numbers of explanatory variables; a
variable is only ”worth adding” if it raises R̄2 .

5 Testing the Overall Significance of the Model: The F-Test

The F-test asks: do the explanatory variables, taken together, explain a significant proportion of the
variation in Y ?

H0 : β1 = β2 = · · · = βk = 0 H1 : at least one βj ̸= 0

SSR/k R2 /k
F = = ∼ F(k, n−k−1)
SSE/(n − k − 1) (1 − R2 )/(n − k − 1)
Decision rule: Reject H0 if Fcalc > Fcritical (from F-tables at the chosen significance level,
with k and n − k − 1 degrees of freedom), or if the associated p-value < α.

6 ANOVA (Analysis of Variance) Table

Regression results are conventionally summarised in an ANOVA table:

Source SS df Mean Square F

Regression SSR k M SR = SSR/k M SR/M SE
Residual (Error) SSE n−k−1 M SE = SSE/(n − k − 1)
Total SST n−1

7 Variance of the Error Term

SSE
σ̂ 2 = M SE =
n−k−1
The divisor is n − k − 1 because
√ k + 1 parameters (β0 , β1 , . . . , βk ) have been estimated, using up
k + 1 degrees of freedom. s = σ̂ 2 is the standard error of the regression.

8 Testing Individual Coefficients: The t-Test

Each coefficient can be tested separately for significance, holding the others fixed:
H0 : βj = 0 H1 : βj ̸= 0

bj − 0
tcalc = ∼ t(n−k−1)
se(bj )
Reject H0 if |tcalc | > tcritical at (n − k − 1) degrees of freedom, meaning Xj is a statistically
significant predictor of Y given the other regressors in the model.

For k = 2, the standard errors of the slope estimates are:

s s
σ̂ 2 x22 σ̂ 2 x21
P P
se(b1 ) = se(b2 ) =
( x21 )( x22 ) − ( x1 x2 )2 ( x21 )( x22 ) − ( x1 x2 )2
P P P P P P

4
Multiple Linear Regression Lecture Notes

9 Multicollinearity

Multicollinearity occurs when two or more explanatory variables are highly correlated with each
other.

Consequences of (severe) multicollinearity

• Coefficient estimates remain unbiased, but their standard errors become large (im-
precise estimates).
• t-statistics shrink, so individually significant variables may appear insignificant.
• Coefficients become very sensitive to small changes in the data or model specification.
• R2 can remain high even though individual coefficients are unreliable — this contradic-
tion (high R2 , insignificant t’s) is a classic diagnostic sign.

Detecting multicollinearity

• Simple correlation coefficient between explanatory variables, rX1 X2 : a value close to ±1

signals a problem.
• Variance Inflation Factor (VIF):
1
V IFj =
1 − Rj2

where Rj2 is the R2 from regressing Xj on all other explanatory variables. A common rule
of thumb: V IFj > 10 indicates serious multicollinearity.

Remedies: drop a redundant variable, collect more data, combine correlated variables into an
index, or use techniques such as ridge regression.

5
Multiple Linear Regression Lecture Notes

10 Example 1

Example 1: Sales Explained by Advertising and Sales Staff

A firm records monthly sales (Y , $’000), advertising expenditure (X1 , $’000), and number of
salespeople (X2 ) over 6 months:

X1 5 6 8 9 10 12
X2 7 4 6 3 8 5
Y 21 19 25 22 31 27

Required: (i) the OLS regression equation; (ii) R2 and adjusted R2 ; (iii) the variance of the
error term; (iv) test the overall significance of the model at 5%; (v) test the significance of each
slope coefficient at 5%.
Step 1: Means.
n = 6, X̄1 = 8.333, X̄2 = 5.5, Ȳ = 24.167
Step 2: Deviation sums (computed from a deviations table, as in simple regression, but now
for three pairs of variables):
X X X
x21 = 33.333 x22 = 17.5 x1 x2 = −1.0
X X X
x1 y = 42.667 x2 y = 24.5 y 2 = 96.833
Step 3: Solve the normal equations for b1 , b2 .
Determinant = (33.333)(17.5) − (−1.0)2 = 583.33 − 1 = 582.33
(42.667)(17.5) − (24.5)(−1.0) 746.67 + 24.5
b1 = = = 1.3243
582.33 582.33
(33.333)(24.5) − (−1.0)(42.667) 816.67 + 42.67
b2 = = = 1.4757
582.33 582.33
b0 = Ȳ − b1 X̄1 − b2 X̄2 = 24.167 − (1.3243)(8.333) − (1.4757)(5.5) = 5.015

Ŷ = 5.015 + 1.3243X1 + 1.4757X2

Step 4: Goodness of fit.
X X
SSR = (Ŷi − Ȳ )2 = 92.656 SSE = y 2 − SSR = 96.833 − 92.656 = 4.177
SSR 92.656
R2 = = = 0.9569 (95.69%)
 SST 96.833

6−1
R̄2 = 1 − (1 − 0.9569) = 1 − (0.0431)(1.667) = 0.9281 (92.81%)
6−2−1
Step 5: Variance of the error term.
SSE
σ̂ 2 = =
n−k−1
Step 6: F-test for overall significance.
SSR/k
Fcalc =
SSE/(n − k − 1)
Step 7: t-tests for individual coefficients.
r
(1.3924)(17.5) 1.3243
se(b1 ) = = 0.2046 tb1 = = 6.474
582.33 0.2046
r
(1.3924)(33.333) 1.4757
se(b2 ) = = 0.2823 tb2 = = 5.227
582.33 0.2823
Conclusion:

6
Multiple Linear Regression Lecture Notes

11 Example 1 — Matrix Approach

Verifying Example 1 Using Matrix Algebra

The same problem can be solved using β̂ = (X′ X)−1 X′ Y. With

  
1 5 7 21
1 6 4  19
   
1 8 6  25
X=
 , Y= 
1 9 3 
 22
 
1 10 8 31
1 12 5 27

we obtain:    
6 50 33 145
X′ X = 50 450 274 X′ Y = 1251
33 274 199 822
Inverting X′ X and multiplying by X′ Y gives:
 
5.015
β̂ = (X′ X)−1 X′ Y = 1.3243
1.4757

This exactly matches b0 , b1 , b2 obtained from the deviation-form normal equations in Example 1
above — the two methods are algebraically identical. In an exam, use whichever method you
are more comfortable computing quickly and accurately; for k = 2, the deviation-form method
is almost always faster by hand.

7
Multiple Linear Regression Lecture Notes

12 Example 2

Example 2: Examination Score Explained by Study Hours and Attendance

A lecturer records exam score (Y , out of 100), weekly study hours (X1 ), and attendance rate
(X2 , %) for 7 students:

X1 2 3 4 5 6 7 8
X2 70 60 80 65 90 72 85
Y 52 50 61 58 70 66 75

Required: (i) the OLS regression equation; (ii) R2 and adjusted R2 ; (iii) the variance of the
error term; (iv) test the overall significance of the model at 5%; (v) test the significance of each
slope coefficient at 5%.

8
Multiple Linear Regression Lecture Notes

13 Example 3: Detecting Multicollinearity

Example 3: A Case of Severe Multicollinearity

Reconsider a firm’s sales (Y ) regressed on advertising (X1 ) and a second promotional variable,
“promotional visits” (X2 ), which happens to move very closely with advertising spend:

X1 5 6 8 9 10 12
X2 3 4 4 5 6 7
Y 18 21 26 30 34 40

Following the same deviation-form procedure as Example 1:

X X X
x21 = 33.333, x22 = 10.833, x1 x2 = 18.333

b1 = 2.400, b2 = 1.400, b0 = 1.400

Ŷ = 1.400 + 2.400X1 + 1.400X2 , R2 = 0.9988, R̄2 = 0.9980

Step: Check correlation between X1 and X2 .
P
x1 x2 18.333 18.333
rX1 X2 = pP 2 P 2 = p = = 0.9648
x1 x2 (33.333)(10.833) 19.00

Step: Compute the Variance Inflation Factor.

1 1 1
V IF = 2 = = = 14.44
1 − rX1 X2 1 − 0.9308 0.0692

Diagnosis: rX1 X2 = 0.96 and V IF = 14.44 ≫ 10 indicate severe multicollinearity

between advertising and promotional visits — the two variables move almost together, so
the model cannot reliably separate their individual effects on sales. Although R2 is very high
(0.9988), this reflects the strong joint explanatory power of the two variables, not necessarily
precise individual coefficient estimates. In practice, a researcher would consider dropping one
variable, combining both into a single ”promotional intensity” index, or collecting additional
data with more independent variation in X1 and X2 .

14 Summary of Key Formulas

9
Multiple Linear Regression Lecture Notes

Quantity Formula

Model Yi = β0 + β1 X1i + · · · + βk Xki + εi

OLS estimator (matrix) β̂ = (X′ X)−1 X′ Y

( x1 y)( x22 ) − ( x2 y)( x1 x2 )
P P P P
Slopes (k = 2, deviation form) b1 = (similarly
( x21 )( x22 ) − ( x1 x2 )2
P P P
for b2 )

Intercept b0 = Ȳ − b1 X̄1 − b2 X̄2

SST = (Yi − Ȳ )2
P
Total sum of squares

SSR = (Ŷi − Ȳ )2
P
Regression sum of squares

Error sum of squares SSE = SST − SSR

Coefficient of determination R2 = SSR/SST = 1 − SSE/SST

 
n−1
Adjusted R2 2 2
R̄ = 1 − (1 − R )
n−k−1
Variance of error term σ̂ 2 = SSE/(n − k − 1)
SSR/k
F-statistic (overall significance) F = ∼ F(k,n−k−1)
SSE/(n − k − 1)
bj
t-statistic (individual coeffi- t= ∼ t(n−k−1)
se(bj )
cient)

Variance Inflation Factor V IFj = 1/(1 − Rj2 )

10
Multiple Linear Regression Lecture Notes

15 Past Paper Style Questions

Question 1 (Past Paper Style — 20 marks)

A retail chain wishes to explain monthly store profit (Y , $’000) using floor space (X1 , ’000 sq.
ft) and number of staff (X2 ) across 6 stores:

X1 2 3 3 4 5 6
X2 8 10 12 11 15 16
Y 30 38 40 44 55 62

Required:

(i) Estimate the multiple regression equation of Y on X1 and X2 . (7 marks)

(ii) Compute R2 and the adjusted R2 , and interpret both. (5 marks)

(iii) Compute the variance of the error term. (3 marks)

(iv) Test the overall significance of the regression at the 5% level. (3 marks)

(v) Briefly explain one limitation of relying on R2 alone to judge model quality. (2 marks)

Question 2 (Past Paper Style — 15 marks)

An economist estimates the following multiple regression relating household consumption ex-
penditure (Y ) to income (X1 ) and household size (X2 ), based on a sample of 20 households:

Ŷ = 250 + 0.65X1 + 120X2 , R2 = 0.82, n = 20, k = 2

Additional information: se(b1 ) = 0.12, se(b2 ) = 45.0, SST = 900,000.

Required:

(i) Interpret the coefficients b1 and b2 . (4 marks)

(ii) Compute the adjusted R2 . (3 marks)

(iii) Test, at the 5% significance level, whether income and household size are each individually
significant determinants of consumption. (5 marks)

(iv) Compute SSE and the variance of the error term. (3 marks)

Question 3 (Past Paper Style — Short Answer, 10 marks)

(a) Explain the difference between R2 and adjusted R2 in a multiple regression context, and
state why the adjusted measure is preferred when comparing models with different num-
bers of regressors. (3
marks)

(b) Define multicollinearity and state two of its consequences for OLS coefficient estimates.
(3 marks)

(c) Explain how the Variance Inflation Factor (VIF) is used to detect multicollinearity, in-
cluding the typical threshold value used as a rule of thumb. (2
marks)

(d) Explain why the divisor used to estimate the variance of the error term in a multiple
regression with k explanatory variables is (n − k − 1) rather than (n − 1) or n. (2 marks)

11
Multiple Linear Regression Lecture Notes

A Appendix: Statistical Tables

A.1 Student’s t-Distribution: Critical Values

Use this table for the individual-coefficient t-test. Find the row matching your degrees of freedom,
df = n − k − 1, and the column matching your (two-tailed) significance level α.

df α = 0.10 α = 0.05 α = 0.02 α = 0.01

(10%) (5%) (2%) (1%)
1 6.314 12.706 31.821 63.657
2 2.920 4.303 6.965 9.925
3 2.353 3.182 4.541 5.841
4 2.132 2.776 3.747 4.604
5 2.015 2.571 3.365 4.032
6 1.943 2.447 3.143 3.707
7 1.895 2.365 2.998 3.499
8 1.860 2.306 2.896 3.355
9 1.833 2.262 2.821 3.250
10 1.812 2.228 2.764 3.169
11 1.796 2.201 2.718 3.106
12 1.782 2.179 2.681 3.055
13 1.771 2.160 2.650 3.012
14 1.761 2.145 2.624 2.977
15 1.753 2.131 2.602 2.947
16 1.746 2.120 2.583 2.921
17 1.740 2.110 2.567 2.898
18 1.734 2.101 2.552 2.878
19 1.729 2.093 2.539 2.861
20 1.725 2.086 2.528 2.845
21 1.721 2.080 2.518 2.831
22 1.717 2.074 2.508 2.819
23 1.714 2.069 2.500 2.807
24 1.711 2.064 2.492 2.797
25 1.708 2.060 2.485 2.787
26 1.706 2.056 2.479 2.779
27 1.703 2.052 2.473 2.771
28 1.701 2.048 2.467 2.763
29 1.699 2.045 2.462 2.756
30 1.697 2.042 2.457 2.750
40 1.684 2.021 2.423 2.704
60 1.671 2.000 2.390 2.660
120 1.658 1.980 2.358 2.617
∞ 1.645 1.960 2.326 2.576

Example use: for Worked Example 1 (n = 6, k = 2), df = 3 and α = 0.05 (two-tailed) ⇒ tcritical =
3.182, matching the value used in the worked solution.

15
Multiple Linear Regression Lecture Notes

A.2 F-Distribution: Critical Values at α = 0.05

Use this table for the overall F-test. Find the column matching df1 = k (numerator, number of
explanatory variables) and the row matching df2 = n − k − 1 (denominator).

df2 \df1 1 2 3 4 5 6 7 8

1 161.45 199.50 215.71 224.58 230.16 233.99 236.77 238.88

2 18.51 19.00 19.16 19.25 19.30 19.33 19.35 19.37
3 10.13 9.55 9.28 9.12 9.01 8.94 8.89 8.85
4 7.71 6.94 6.59 6.39 6.26 6.16 6.09 6.04
5 6.61 5.79 5.41 5.19 5.05 4.95 4.88 4.82
6 5.99 5.14 4.76 4.53 4.39 4.28 4.21 4.15
7 5.59 4.74 4.35 4.12 3.97 3.87 3.79 3.73
8 5.32 4.46 4.07 3.84 3.69 3.58 3.50 3.44
9 5.12 4.26 3.86 3.63 3.48 3.37 3.29 3.23
10 4.96 4.10 3.71 3.48 3.33 3.22 3.14 3.07
11 4.84 3.98 3.59 3.36 3.20 3.09 3.01 2.95
12 4.75 3.89 3.49 3.26 3.11 3.00 2.91 2.85
13 4.67 3.81 3.41 3.18 3.03 2.92 2.83 2.77
14 4.60 3.74 3.34 3.11 2.96 2.85 2.76 2.70
15 4.54 3.68 3.29 3.06 2.90 2.79 2.71 2.64
16 4.49 3.63 3.24 3.01 2.85 2.74 2.66 2.59
17 4.45 3.59 3.20 2.96 2.81 2.70 2.61 2.55
18 4.41 3.55 3.16 2.93 2.77 2.66 2.58 2.51
19 4.38 3.52 3.13 2.90 2.74 2.63 2.54 2.48
20 4.35 3.49 3.10 2.87 2.71 2.60 2.51 2.45
21 4.32 3.47 3.07 2.84 2.68 2.57 2.49 2.42
22 4.30 3.44 3.05 2.82 2.66 2.55 2.46 2.40
23 4.28 3.42 3.03 2.80 2.64 2.53 2.44 2.37
24 4.26 3.40 3.01 2.78 2.62 2.51 2.42 2.36
25 4.24 3.39 2.99 2.76 2.60 2.49 2.40 2.34
26 4.23 3.37 2.98 2.74 2.59 2.47 2.39 2.32
27 4.21 3.35 2.96 2.73 2.57 2.46 2.37 2.31
28 4.20 3.34 2.95 2.71 2.56 2.45 2.36 2.29
29 4.18 3.33 2.93 2.70 2.55 2.43 2.35 2.28
30 4.17 3.32 2.92 2.69 2.53 2.42 2.33 2.27
40 4.08 3.23 2.84 2.61 2.45 2.34 2.25 2.18
60 4.00 3.15 2.76 2.53 2.37 2.25 2.17 2.10
120 3.92 3.07 2.68 2.45 2.29 2.18 2.09 2.02

Example use: for Worked Example 1 (k = 2, df2 = 3), Fcritical = 9.55; for Worked Example 2
(k = 2, df2 = 4), Fcritical = 6.94 — both match the values used in the worked solutions.

16
Multiple Linear Regression Lecture Notes

A.3 F-Distribution: Critical Values at α = 0.01

df2 \df1 1 2 3 4 5 6 7 8

1 4052.18 4999.50 5403.35 5624.58 5763.65 5858.99 5928.36 5981.07

2 98.50 99.00 99.17 99.25 99.30 99.33 99.36 99.37
3 34.12 30.82 29.46 28.71 28.24 27.91 27.67 27.49
4 21.20 18.00 16.69 15.98 15.52 15.21 14.98 14.80
5 16.26 13.27 12.06 11.39 10.97 10.67 10.46 10.29
6 13.75 10.92 9.78 9.15 8.75 8.47 8.26 8.10
7 12.25 9.55 8.45 7.85 7.46 7.19 6.99 6.84
8 11.26 8.65 7.59 7.01 6.63 6.37 6.18 6.03
9 10.56 8.02 6.99 6.42 6.06 5.80 5.61 5.47
10 10.04 7.56 6.55 5.99 5.64 5.39 5.20 5.06
11 9.65 7.21 6.22 5.67 5.32 5.07 4.89 4.74
12 9.33 6.93 5.95 5.41 5.06 4.82 4.64 4.50
13 9.07 6.70 5.74 5.21 4.86 4.62 4.44 4.30
14 8.86 6.51 5.56 5.04 4.69 4.46 4.28 4.14
15 8.68 6.36 5.42 4.89 4.56 4.32 4.14 4.00
16 8.53 6.23 5.29 4.77 4.44 4.20 4.03 3.89
17 8.40 6.11 5.18 4.67 4.34 4.10 3.93 3.79
18 8.29 6.01 5.09 4.58 4.25 4.01 3.84 3.71
19 8.18 5.93 5.01 4.50 4.17 3.94 3.77 3.63
20 8.10 5.85 4.94 4.43 4.10 3.87 3.70 3.56
21 8.02 5.78 4.87 4.37 4.04 3.81 3.64 3.51
22 7.95 5.72 4.82 4.31 3.99 3.76 3.59 3.45
23 7.88 5.66 4.76 4.26 3.94 3.71 3.54 3.41
24 7.82 5.61 4.72 4.22 3.90 3.67 3.50 3.36
25 7.77 5.57 4.68 4.18 3.85 3.63 3.46 3.32
26 7.72 5.53 4.64 4.14 3.82 3.59 3.42 3.29
27 7.68 5.49 4.60 4.11 3.78 3.56 3.39 3.26
28 7.64 5.45 4.57 4.07 3.75 3.53 3.36 3.23
29 7.60 5.42 4.54 4.04 3.73 3.50 3.33 3.20
30 7.56 5.39 4.51 4.02 3.70 3.47 3.30 3.17
40 7.31 5.18 4.31 3.83 3.51 3.29 3.12 2.99
60 7.08 4.98 4.13 3.65 3.34 3.12 2.95 2.82
120 6.85 4.79 3.95 3.48 3.17 2.96 2.79 2.66

Reading tip: If your exam paper supplies its own statistical tables, always use those values
instead — minor differences can arise between table editions due to rounding. These tables are
provided as a self-contained reference matching the worked examples in these notes.

17