TABLE 3.

2: HYPOTHETICAL DATA ON WEEKLY FAMILY CONSUMPTION EXPENDITURE Y AND WEEKLY FAMILY IN

Source: Gujarati, D.N. (2004) Basic Econometrics.
Sample Data Calculations P
Numerator Denominator of b2_hat
Y X Yi - Y-bar Xi - X-bar (Xi - X-bar) (Xi - X-bar)^2
Y X yi xi yixi x-sqr
70 80 -41 -90 3690 8100
65 100 -46 -70 3220 4900
90 120 -21 -50 1050 2500
95 140 -16 -30 480 900
110 160 -1 -10 10 100
115 180 4 10 40 100
120 200 9 30 270 900
140 220 29 50 1450 2500
155 240 44 70 3080 4900
150 260 39 90 3510 8100
Mean 111 170
Sum 0 0 16800 33000
n 10
k 2
df 8
Correlation

Estimates Experiment
b2_hat b1_hat b2_hat b1_hat
Estimate (Method 1) 0.5091 24.4545 0.5 26
Estimate (Method 2) 0.5091
Estimate (Method 3) 0.5091
Sigma_hat-sqr 42.16
Var (sigma) 0.0013 41.1371
se 0.0357 6.4138 Regression Line
Cov (b1_hat, b2_hat) -0.2172 -0.2172 Yi_hat = 24.45 + 0.5091*Xi

r-sqr ESS/TSS 0.9621

r-sqr (Alternative 1) 1 - (RSS/TS 0.9621
r-sqr (Alternative 2) b2_hat-sqr 0.9621
r (Correlation) 0.9808

Properties of Regression Line (Verify Each)

1. It passes through the sample means of 𝑌 and 𝑋.
2. The mean value of the estimated 𝑌 is equal to the mean value of the actual 𝑌.
3. The mean value of the residuals 𝑢 ̂_𝑖 is zero
4. The residuals 𝑢 ̂_𝑖 are uncorrelated with the predicted 𝑌_ 𝑖.
5. The residuals 𝑢 ̂_𝑖 are uncorrelated with 𝑋_𝑖.
Confidence Interval Estimation
Null Hyp 0.3
alpha 0.05
t_alpha/2
Lower Confidence
Upper Confidence
Decision Rule Since 0.3 does not fall within (0.4267, 0.5915), we reject the null hypthesis.

Hypothesis Testing (Two-Sided Test)

Null Hyp 0.3
t-stat - computed
t-stat - critical value
Lower Confidence
Upper Confidence

Decision Rule
Reject Null No 5.85 > 2.31
Decision Rule Since 5.85 > 2.305, we reject the null hypthesis at 5% level of significance.
Decision Rule - Alt Since 0.5091 lies outside (0.2176, 0.3824) (i.e., critical regions) we reject the null hypothesis.

Hypothesis Testing (One-Sided Test)

t-stat - critical value 1.86
Upper Confidence 0.3665

Decision Rule Decision

Is computed t-stat > t cNo Reject the Null
Is estimated beta > hy Yes Reject the Null
E Y AND WEEKLY FAMILY INCOME X

Calculations Parameter Estimation

Denominator of b2_hat
(Xi - X-bar)^2 (Yi - Y-bar)^2
y-sqr X-sqr YX yi_hat ui_hat ui_hat^2
1681 6400 5600 65.18 4.82 23.21
2116 10000 6500 75.36 -10.36 107.40
441 14400 10800 85.55 4.45 19.84
256 19600 13300 95.73 -0.73 0.53
1 25600 17600 105.91 4.09 16.74
16 32400 20700 116.09 -1.09 1.19
81 40000 24000 126.27 -6.27 39.35
841 48400 30800 136.45 3.55 12.57
1936 57600 37200 146.64 8.36 69.95
1521 67600 39000 156.82 -6.82 46.49
111
8890 322000 205500 0.00 337.27

170

150

130

110

90

70

50
50 70 90 110 130 150 170 190 210 230 250 270 290

Y yi_hat Linear (yi_hat)

eject the null hypothesis.
 Verifying Assumptions Experiment

ui_hat ui_hat * yi_hat ui_hat * Xi yi_hat ui_hat ui_hat^2

4.82 314.06 385.45 66 4 16.00
-10.36 -781.04 -1036.36 76 -11 121.00
4.45 381.07 534.55 86 4 16.00
-0.73 -69.62 -101.82 96 -1 1.00
4.09 433.26 654.55 106 4 16.00
-1.09 -126.64 -196.36 116 -1 1.00
-6.27 -792.07 -1254.55 126 -6 36.00
3.55 483.79 780.00 136 4 16.00
8.36 1226.41 2007.27 146 9 81.00
-6.82 -1069.21 -1772.73 156 -6 36.00

0.00 0.00 0.00 0.00 340.00

1.037526E-15

270 290
 Goodness of Fit

TSS ESS RSS

Yi –Y-bar)^Yi_hat –Y-bui_hat^2
1681 2099.31 23.21
2116 1269.95 107.40
441 647.93 19.84
256 233.26 0.53
1 25.92 16.74
16 25.92 1.19
81 233.26 39.35
841 647.93 12.57
1936 1269.95 69.95
1521 2099.31 46.49

8890.00 8552.73 337.27 8890.00

TABLE 3.2: HYPOTHETICAL DATA ON WEEKLY FAMILY CONSUMPTION EXPENDITURE Y AND WEEKLY FAMILY IN
Source: Gujarati, D.N. (2004) Basic Econometrics.
Sample Data Calculations P
Numerator Denominator of b2_hat
Y X Yi - Y-bar Xi - X-bar (Xi - X-bar) (Xi - X-bar)^2
Y X yi xi yixi x-sqr
70 80 -41 -90 3690 8100
65 100 -46 -70 3220 4900
90 120 -21 -50 1050 2500
95 140 -16 -30 480 900
110 160 -1 -10 10 100
115 180 4 10 40 100
120 200 9 30 270 900
140 220 29 50 1450 2500
155 240 44 70 3080 4900
150 260 39 90 3510 8100
Mean 111 170
Sum 0 0 16800 33000
n 10
k 2
df 8
Correlation

Estimates Experiment
b2_hat b1_hat b2_hat b1_hat
Estimate (Method 1) 0.5091 24.4545 0.5 26
Estimate (Method 2) 0.5091
Estimate (Method 3) 0.5091
Sigma_hat-sqr 42.16
Var (sigma) 0.0013 41.1371
se 0.0357 6.4138 Regression Line
Cov (b1_hat, b2_hat) -0.2172 -0.2172 Yi_hat = 24.45 + 0.5091*Xi

r-sqr ESS/TSS 0.9621

r-sqr (Alternative 1) 1 - (RSS/TS 0.9621
r-sqr (Alternative 2) b2_hat-sqr 0.9621
r (Correlation) 0.9808

Properties of Regression Line (Verify Each)

1. It passes through the sample means of 𝑌 and 𝑋.
2. The mean value of the estimated 𝑌 is equal to the mean value of the actual 𝑌.
3. The mean value of the residuals 𝑢 ̂_𝑖 is zero
4. The residuals 𝑢 ̂_𝑖 are uncorrelated with the predicted 𝑌_ 𝑖.
5. The residuals 𝑢 ̂_𝑖 are uncorrelated with 𝑋_𝑖.
Confidence Interval Estimation
Null Hyp 0.3
alpha 0.05
t_alpha/2 2.3060
Lower Confidence 0.4267 9.6643
Upper Confidence 0.5915 39.2448
Decision Rule Since 0.3 does not fall within (0.4267, 0.5915), we reject the null hypthesis.

Hypothesis Testing (Two-Sided Test)

Null Hyp 0.3
t-stat - computed 5.85
t-stat - critical value 2.31
Lower Confidence 0.2176
Upper Confidence 0.3824

Decision Rule
Reject Null Yes 5.85 > 2.31
Decision Rule Since 5.85 > 2.305, we reject the null hypthesis at 5% level of significance.
Decision Rule - Alt Since 0.5091 lies outside (0.2176, 0.3824) (i.e., critical regions) we reject the null hypothesis.

Hypothesis Testing (One-Sided Test)

t-stat - critical value 1.86
Upper Confidence 0.3665

Decision Rule Decision

Is computed t-stat > t cYes Reject the Null
Is estimated beta > hy Yes Reject the Null
E Y AND WEEKLY FAMILY INCOME X

Calculations Parameter Estimation

Denominator of b2_hat
(Xi - X-bar)^2 (Yi - Y-bar)^2
y-sqr X-sqr YX yi_hat ui_hat ui_hat^2
1681 6400 5600 65.18 4.82 23.21
2116 10000 6500 75.36 -10.36 107.40
441 14400 10800 85.55 4.45 19.84
256 19600 13300 95.73 -0.73 0.53
1 25600 17600 105.91 4.09 16.74
16 32400 20700 116.09 -1.09 1.19
81 40000 24000 126.27 -6.27 39.35
841 48400 30800 136.45 3.55 12.57
1936 57600 37200 146.64 8.36 69.95
1521 67600 39000 156.82 -6.82 46.49
111
8890 322000 205500 0.00 337.27

170

150

130

110

90

70

50
50 70 90 110 130 150 170 190 210 230 250 270 290

Y yi_hat Linear (yi_hat)

eject the null hypothesis.
 Verifying Assumptions Experiment

ui_hat ui_hat * yi_hat ui_hat * Xi yi_hat ui_hat ui_hat^2

4.82 314.06 385.45 66 4 16.00
-10.36 -781.04 -1036.36 76 -11 121.00
4.45 381.07 534.55 86 4 16.00
-0.73 -69.62 -101.82 96 -1 1.00
4.09 433.26 654.55 106 4 16.00
-1.09 -126.64 -196.36 116 -1 1.00
-6.27 -792.07 -1254.55 126 -6 36.00
3.55 483.79 780.00 136 4 16.00
8.36 1226.41 2007.27 146 9 81.00
-6.82 -1069.21 -1772.73 156 -6 36.00

0.00 0.00 0.00 0.00 340.00

1.037526E-15

270 290
 Goodness of Fit

TSS ESS RSS

Yi –Y-bar)^Yi_hat –Y-bui_hat^2
1681 2099.31 23.21
2116 1269.95 107.40
441 647.93 19.84
256 233.26 0.53
1 25.92 16.74
16 25.92 1.19
81 233.26 39.35
841 647.93 12.57
1936 1269.95 69.95
1521 2099.31 46.49

8890.00 8552.73 337.27 8890.00