UNIT 4:

Data:

1.ESTIMATION OF LOG-LINEAR MODEL:

Y = aXb

lnY = lna + blnX

Y* = A + bX*

Where lnY=Y* lnX=X* lna=A

REPORTING THE RESULTS:

 Y = aXb

lnY = lna + blnX

Y* = A + bX*

Coefficient: 0.851030 0.644997

Standard Error: 5.770538 1.190858

t-Statistic: 0.147478 0.541624

Prob.: 0.8857 0.5999

Confidence Interval (95%) Low -12.00653 -2.008400

High 13.70859 3.298395

R-squared: 0.028500 Adjusted R-squared: -0.068650

F-statistic: 0.293356 Prob(F-statistic): 0.599936

CONFIDENCE INTERVALS:
INTERPRETATION THE RESULTS:
The results for your log-linear regression model can be interpreted as follows:

1. Model Equation: The log-linear regression model is given by the equation ln(Y) = 0.851030 + 0.644997
* ln(X), where ln(Y) is the natural logarithm of the dependent variable Y, and ln(X) is the natural
logarithm of the independent variable X.

2. Coefficients:

- The coefficient of ln(X) is 0.644997. This means that for a 1% increase in ln(X), ln(Y) is expected to
increase by approximately 0.644997%.

- The coefficient of the intercept term A is 0.851030. This represents the expected value of ln(Y) when
ln(X) is zero.

3. Standard Errors:

- The standard error of the coefficient for ln(X) is 1.190858.

- The standard error of the intercept term A is 5.770538.

4. T-Statistics and P-Values:

- The t-statistic for ln(X) is 0.541624 with a p-value of 0.5999. This t-statistic measures whether the
coefficient for ln(X) is statistically significant. In this case, with a p-value of 0.5999, it is not statistically
significant at conventional significance levels (e.g., 0.05).

- The t-statistic for the intercept term A is 0.147478 with a p-value of 0.8857. Like ln(X), the intercept
term is also not statistically significant.

5. R-Squared and Adjusted R-Squared:

- The R-squared value is 0.028500. This indicates that only 2.85% of the variance in ln(Y) can be
explained by the model, suggesting that the model does not fit the data well.

- The adjusted R-squared is -0.06850, which is negative. This is unusual and might indicate that the
model is a poor fit for the data.

6. F-Statistic and P-Value:

- The F-statistic is 0.293356 with a p-value of 0.599936. The F-statistic tests the overall significance of
the model. A high p-value suggests that the model as a whole is not statistically significant, which is the
case here.

7. Confidence Intervals:

- The 95% confidence interval for the coefficient of ln(X) is approximately -3.30 to -2.01. This means
that we are 95% confident that the true value of the coefficient lies within this range.

- The 95% confidence interval for the intercept term A is approximately -12.01 to 13.71. This wide
interval indicates a lack of precision in estimating the intercept.
In summary, the log-linear regression model does not seem to be a good fit for the data. Both the
coefficients and the intercept term are not statistically significant, and the model's explanatory power
(R-squared and adjusted R-squared) is very low. Additionally, the F-statistic suggests that the model is
not significant. You may want to consider a different model or investigate further to understand the
relationship between ln(Y) and ln(X) better.

2.ESTIMATION OF LOG-LIN (SEMI LOG MODEL):

Y = Y0 (1+r)X

lnY = lnY0 + Xln(1+r)

lnY = B1 + B2X

Where, lnY0= B1 ln(1+r)=B2

REPORTING THE RESULTS:

Y = Y0 (1+r)X

lnY = lnY0 + Xln(1+r)

lnY = B1 + B2X

Coefficient: 3.333568 0.005014

Standard Error: 1.226364 0.009507

t-Statistic: 2.718253 0.527432

Prob.: 0.0216 0.6094

 Confidence Interval (95%) Low 0.601059 -0.016169

High 6.066077 0.026198

R-squared: 0.027065 Adjusted R-squared: -0.070228

F-statistic: 0.278184 Prob(F-statistic): 0.609396

CONFIDENCE INTERVALS:

INTERPRETATION THE RESULTS:

The provided information is related to a simple regression model in log-linear form:

ln(Y) = A + B * ln(X)

Here are the key statistics and their interpretations:

1. Coefficients:

- Coefficient of A (intercept): 3.333568

- Coefficient of B (slope, related to ln(X)): 0.0050140

Interpretation: A represents the intercept, and B represents the coefficient associated with the natural
logarithm of X. In this context, A represents the value of ln(Y) when ln(X) is 0. B represents the change in
ln(Y) for a one-unit change in ln(X).

2. Standard Errors:

- Standard error of A: 1.226364

- Standard error of B: 0.009507

 Interpretation: These values indicate the precision of the estimated coefficients. Smaller standard
errors suggest more reliable estimates.

3. t-Statistics:

- t statistic of A: 2.718253

- t statistic of B: 0.527432

Interpretation: These t-statistics measure the significance of the coefficients. A high t-statistic (in
absolute value) suggests that the coefficient is significant, while a low t-statistic indicates that the
coefficient may not be significant.

4. p-Values:

- p-value of A: 0.0216

- p-value of B: 0.6094

Interpretation: The p-values are associated with the null hypothesis that the corresponding coefficient
is equal to zero. A low p-value (typically below 0.05) suggests that the coefficient is statistically
significant. In this case, the coefficient of A is statistically significant, but the coefficient of B is not.

5. R-squared: 0.027065

Adjusted R-squared: -0.070228

Interpretation: The R-squared value (and adjusted R-squared) measures the goodness of fit of the
model. In this case, R-squared is quite low, indicating that the model does not explain much of the
variance in the data. The adjusted R-squared considers the number of independent variables in the
model and penalizes for adding unnecessary variables. A negative adjusted R-squared may indicate that
the model is not a good fit.

6. F-Statistic: 0.278184

p-value for F-Statistic: 0.609396

Interpretation: The F-statistic measures the overall significance of the regression model. A low F-
statistic and a high p-value for the F-statistic suggest that the model may not be a good fit for the data.

7. Confidence Intervals:

- Confidence intervals for X: [low = -0.016169, high = 0.026198]

- Confidence intervals for A: [low = 0.601059, high = 6.066077]

Interpretation: These intervals provide a range within which the true coefficients are likely to lie. For X,
the confidence interval includes zero, indicating that the coefficient is not statistically significant. For A,
the confidence interval is relatively wide, indicating some uncertainty in its estimation.

In summary, the log-linear regression model suggests that the coefficient of A is statistically significant,
but the coefficient of B is not. The low R-squared and adjusted R-squared values indicate that the model
does not explain much of the variation in the data. The F-statistic and its associated p-value also suggest
that the overall model may not be a good fit for the data.

3.ESTIMATION OF LIN-LOG (SEMI LOG MODEL):

Y = A + BlnX

REPORTING THE RESULTS:

Y = A + BlnX

Coefficient: -87.97060 30.32343

Standard Error: 271.4137 56.01130

t-Statistic: -0.324120 0.541381

Prob.: 0.7525 0.6001

Confidence Interval (95%) Low 0.601059 -0.016169

High 0.026198 0.026198

R-squared: 0.028475 Adjusted R-squared: -0.068678

F-statistic: 0.293093 Prob(F-statistic): 0.600098

CONFIDENCE INTERVALS:

INTERPRETATION THE RESULTS:

1. Coefficients:

- Coefficient of X (B) is 30.32343: This means that a 1% increase in lnX is associated with an
approximately 30.32% increase in lnY, assuming all other factors remain constant.

- Coefficient of A is -87.97060: This is the intercept term. It represents the expected value of lnY when
lnX is 0. In this context, it's not very meaningful because lnX should never be zero.

2. Standard Errors:

- Standard Error of A is 271.4137: It measures the uncertainty or variability in the estimated value of A.

- Standard Error of X is 56.01130: It measures the uncertainty or variability in the estimated value of B.

3. t-Statistics:

- t statistic of A is -0.324120: This statistic measures how many standard errors A is away from zero. In
this case, the t statistic is small, suggesting that A is not significantly different from zero.

- t statistic of X is 0.541381: Similarly, this statistic measures how many standard errors B is away from
zero. The t statistic for X is also small, indicating that B is not significantly different from zero.

4. P-values:

- P-value of A is 0.7525: A high p-value for A suggests that A is not statistically significant in explaining
the variation in lnY.

- P-value of X is 0.6001: Similarly, a high p-value for X suggests that B is not statistically significant in
explaining the variation in lnY.
5. R-squared and Adjusted R-squared:

- R-squared is 0.028475: R-squared measures the proportion of the variance in the dependent variable
(lnY) that is explained by the independent variable (lnX). In this case, only about 2.85% of the variance in
lnY is explained by lnX.

- Adjusted R-squared is -0.068678: The adjusted R-squared accounts for the number of predictors in
the model. A negative value suggests that the model doesn't fit the data well.

6. F-Statistic:

- F statistic is 0.293093: The F statistic tests the overall significance of the model. A low F statistic and a
high p-value for the F statistic (0.600098) suggest that the model as a whole is not a good fit for the
data.

7. Confidence Intervals:

- Confidence intervals for X: The interval [-0.016169, 0.026198] represents a range of values for the
coefficient of X, with 95% confidence.

- Confidence intervals for A: The interval [0.601059, 0.026198] for the coefficient of A has an issue, as
the values for the low and high intervals seem to be incorrect. The confidence interval for the intercept
A should not contain negative values, and the high value should not be smaller than the low value.

In summary, based on the provided information, it appears that the model is not a good fit for the data,
as neither A nor B is statistically significant, and the model's explanatory power (R-squared and adjusted
R-squared) is very low. Additionally, there seems to be an issue with the confidence intervals for A.
Further investigation and potentially a different model may be necessary to better understand the
relationship between lnY and lnX.
4.ESTIMATION OF RECIPROCAL MODEL:
Y = A + B(1/X)

REPORTING THE RESULTS:

Y = A + B(1/X)

Coefficient: 89.98887 -3919.247

Standard Error: 55.51615 6933.805

t-Statistic: 1.620950 -0.564372

Prob.: 0.1361 0.5849

Confidence Interval (95%) Low -33.70881 -19362.73

High 213.6866 11536.23

R-squared: 0.030868 Adjusted R-squared: -0.066045

F-statistic: 0.318516 Prob(F-statistic): 0.584936

CONFIDENCE INTERVALS:

INTERPRETATION THE RESULTS:

1. Coefficients:

- Coefficient of A (A): 89.98887

- Coefficient of B (1/X): -3919.247

These coefficients indicate the relationship between the independent variable (X) and the dependent
variable (Y) in your model.

2. Standard Errors:

- Std Error of A: 55.51615

- Std Error of X: 6933.805

These are measures of the variability or uncertainty associated with the respective coefficients.
Smaller standard errors indicate more precise estimates.

3. t-statistics:

- t-statistic of A: 1.620950

- t-statistic of X: -0.564372

The t-statistic measures the number of standard errors that the coefficients are away from zero. In this
case, a t-statistic close to 0 suggests that the coefficient is not significantly different from zero.

4. p-values:

- p-value of A: 0.1361

- p-value of X: 0.5849
 The p-values represent the probability that the coefficient is equal to zero. A small p-value (typically
less than 0.05) suggests that the coefficient is statistically significant. In this case, neither A nor X is
statistically significant at the 0.05 significance level.

5. R-squared:

- R-squared: 0.030868

R-squared measures the proportion of the variance in the dependent variable (Y) that is explained by
the independent variables (A and 1/X). In your model, only about 3.09% of the variance in Y is explained
by the independent variables.

6. Adjusted R-squared:

- Adjusted R-squared: -0.066045

Adjusted R-squared takes into account the number of independent variables in the model and is used
to assess the goodness of fit. A negative adjusted R-squared suggests that the model doesn't fit the data
well.

7. F-statistic:

- F-statistic: -0.066045

The F-statistic tests the overall significance of the model. A negative F-statistic and a high p-value
(0.584936) suggest that the model as a whole is not statistically significant in explaining the variation in
Y.

8. Confidence Intervals:

- Confidence intervals for X: Low = -19362.73, High = 11536.23

- Confidence intervals for A: Low = -33.70881, High = 213.6866

These intervals provide a range within which the true values of the coefficients are likely to lie. In this
case, the confidence interval for X contains zero, indicating that the X coefficient is not statistically
significant.

In summary, the results suggest that neither the intercept (A) nor the coefficient of 1/X (B) is statistically
significant in explaining the variation in the dependent variable Y. The model's low R-squared and
adjusted R-squared values, along with the negative F-statistic, indicate that the model doesn't fit the
data well. You may need to reconsider the model or gather additional data to improve its explanatory
power.
5.ESTIMATION OF LOG RECIPROCAL MODEL:
lnY = A + B(1/X)

REPORTING THE RESULTS:

lnY = A + B(1/X)

Coefficient: 4.621405 -81.35703

Standard Error: 1.181170 147.5247

t-Statistic: 3.912567 -0.551481

Prob.: 0.0029 0.5934

Confidence Interval (95%) Low -33.70881 -19362.73

High 213.6866 11536.23

R-squared: 0.029515 Adjusted R-squared: -0.067533

F-statistic: 0.304131 Prob(F-statistic): 0.593412

CONFIDENCE INTERVALS:

INTERPRETATION THE RESULTS:

1. Coefficients:

- The coefficient of X (B) is -81.35703. This means that for each unit increase in the predictor variable
X, the natural logarithm of the response variable Y is expected to decrease by approximately 81.36 units.
Since it's negative, this suggests an inverse relationship between X and ln(Y).

- The coefficient of the intercept term A is 4.621405. This is the estimated value of ln(Y) when X is zero.

2. Standard Errors:

- The standard error of A is 1.181170, and the standard error of X is 147.5247. These represent the
standard deviation of the coefficient estimates. Smaller standard errors indicate more precise estimates.

3. t-Statistics:

- The t-statistic for A is 3.912567, and the t-statistic for X is -0.551481. These statistics are used to test
the significance of the coefficients. The t-statistic for A is relatively large, indicating that the coefficient A
is likely significant. However, the t-statistic for X is small and negative, suggesting that the coefficient B
(associated with X) is not statistically significant.

4. Probability (p) Values:

- The p-value for A is 0.0029, and the p-value for X is 0.5934. A low p-value (typically below 0.05)
indicates that the corresponding coefficient is statistically significant. In this case, A is statistically
significant (p < 0.05), while X is not statistically significant (p > 0.05).

5. R-squared and Adjusted R-squared:

 - The R-squared value is 0.029515. This represents the proportion of the variance in the dependent
variable Y that is explained by the independent variable X. In this case, X explains only about 2.95% of
the variation in ln(Y).

- The adjusted R-squared is -0.067533. A negative adjusted R-squared indicates that the model may
not be a good fit for the data. It suggests that the model is not providing any meaningful explanation for
the variation in the data.

6. F-Statistic and Probability (p) Value:

- The F-statistic is 0.304131, and the p-value for the F-statistic is 0.593412. The F-statistic is used to
test the overall significance of the model. In this case, the high p-value suggests that the model is not
statistically significant.

7. Confidence Intervals:

- For X, the 95% confidence interval is from -19362.73 to 11536.23. This interval suggests that the true
value of the coefficient B may lie within this range with 95% confidence.

- For A, the 95% confidence interval is from -33.70881 to 213.6866. This interval suggests that the true
value of the intercept A may lie within this range with 95% confidence.

In summary, it appears that the coefficient of A is statistically significant, but the coefficient B
(associated with X) is not. The model doesn't explain much of the variation in the dependent variable,
and the adjusted R-squared is negative, indicating that this model may not be appropriate for your data.
Additionally, the F-statistic and its associated p-value suggest that the model as a whole is not
statistically significant.
UNIT 5: MULTIPLE REGRESSION
Data:

1.ESTIMATION OF LOG-LINEAR MODEL:

Dx = aPxb Mc

lnDx = lna + blnPx + clnM

lnDx = A + blnPx + clnM

REPORTING THE RESULTS:

 Dx = aPxb Mc

lnDx = lna + blnPx + clnM

lnDx = A + blnPx + clnM

Coefficient 4.974534 0.032419 0.477461

Standard Error 0.321129 0.043245 0.054963

t-statistic 15.49077 0.749666 8.686953

prob. value 0.0000 0.4779 0.0001

R- squared: 0.986903 Adjusted R-squared: 0.983160

F-statistic: 263.7270 Prob. (F-statistic): 0.000000

CONFIDENCE INTERVALS:

INTERPRETATION THE RESULTS:

1. Coefficients (A, b, and c):

- A, the intercept, is highly significant with a very low p-value (p=0), indicating that the model's intercept is
significantly different from zero.

- b, the coefficient for ln(P), is not statistically significant as its p-value (p=0.478) is greater than the commonly
used significance level of 0.05. This suggests that the relationship between ln(P) and ln(D) may not be significant.

- c, the coefficient for ln(M), is highly significant with a very low p-value (p=0.0001), indicating that the
relationship between ln(M) and ln(D) is statistically significant.
2. Goodness of Fit:

- The R-squared value (0.987) is very high, indicating that the independent variables in the model (ln(P) and
ln(M)) explain a large proportion of the variance in the dependent variable (ln(D)).

- The adjusted R-squared value (0.983) takes into account the number of predictors and provides a more
conservative measure of model fit.

- The F-statistic (263.727) is very high, and its p-value is 0, indicating that the overall model is statistically
significant. In other words, at least one of the independent variables significantly contributes to explaining the
variance in ln(D).

3. Confidence Intervals:

- The 95% confidence intervals for ln(P) and ln(M) provide a range within which the true coefficients are likely to
lie. For example, the 95% confidence interval for ln(M) is between 4.215 and 5.733.

In summary, this regression model suggests that ln(M) is a statistically significant predictor of ln(D), and the model
is statistically significant. However, ln(P) does not appear to be a significant predictor in this model, as its
coefficient is not statistically different from zero.

2.ESTIMATION OF LOG-LIN (SEMI LOG MODEL):

Dx = Dx0 (1+b)Px (1+c)M

lnDx = lnDx0 + Px. ln(1+b) + M ln(1+c)

lnDx = B1 + B2Px + B3M

REPORTING THE RESULTS:

Dx = Dx0 (1+b)Px (1+c)M

lnDx = lnDx0 + Px. ln(1+b) + M ln(1+c)

 lnDx = B1 + B2Px + B3M

Coefficient 598.1214 -2.448584 9.191987

Standard Error 343.9567 6.293053 5.624100

t-statistic 1.738944 -0.389093 1.634393

prob. value 0.1256 0.7088 0.1462

R- squared: 0.971749 Adjusted R-squared: 0.963678

F-statistic: 120.3911 Prob. (F-statistic): 0.000004

CONFIDENCE INTERVALS:

INTERPRETATION THE RESULTS:

The multiple regression model you've presented, lnD = A + bP + cM, has several statistical parameters and results
that are important for interpretation. Let's go through each of these results one by one:
1. Coefficients:

- A (Intercept) = 598.121

- b (Coefficient for P) = -2.448

- c (Coefficient for M) = 9.192

These coefficients represent the estimated relationship between the natural logarithm of D (lnD) and the predictor
variables P and M.

2. Standard Errors:

- Standard Error for A = 343.957

- Standard Error for b = 6.293

- Standard Error for c = 5.624

Standard errors measure the variability in the coefficient estimates. Smaller standard errors indicate more precise
estimates.

3. t-Statistics:

- t-statistic for A = 1.739

- t-statistic for b = -0.389

- t-statistic for c = 1.634

The t-statistics are used to test the hypothesis that the coefficients are different from zero. In this case, the t-
statistic for A is relatively large, while the t-statistic for b is close to zero, indicating that A may be statistically
significant, but b is not. The t-statistic for c is also relatively large, suggesting that c may be statistically significant.

4. p-Values:

- p-value for A = 0.126

- p-value for b = 0.709

- p-value for c = 0.146

The p-values associated with the coefficients test the null hypothesis that each coefficient is equal to zero. A low p-
value (typically less than 0.05) indicates that the coefficient is statistically significant. In this case, A and c may be
considered statistically significant, while b is not.

5. R-squared and Adjusted R-squared:

- R-squared = 0.972

- Adjusted R-squared = 0.964

R-squared represents the proportion of variance in the dependent variable that is explained by the independent
variables. In this model, a very high R-squared (close to 1) suggests that a significant portion of the variance in lnD
is explained by P and M.

6. F-Statistic:

- F-statistic = 120.391
The F-statistic is used to test whether the overall model is statistically significant. A low p-value for the F-statistic
(in this case, very close to zero) indicates that the model is statistically significant.

7. Confidence Intervals:

- 95% Confidence Intervals for M: Low = 0.347, High = 0.607

- 95% Confidence Intervals for P: Low = -0.027, High = 0.013

- 95% Confidence Intervals for A: Low = 5.287, High = 7.748

The confidence intervals provide a range within which we can be reasonably confident that the true values of the
coefficients lie. In this case, the confidence intervals for M, P, and A do not include zero, which supports their
statistical significance.

In summary, this multiple regression model suggests that variables A and M are statistically significant predictors
of lnD, while variable P is not statistically significant. The model is highly significant, and it explains a substantial
portion of the variance in lnD, as indicated by the high R-squared value. The confidence intervals for the
coefficients provide a range of plausible values for the true coefficients.
3.ESTIMATION OF LIN-LOG (SEMI LOG MODEL):
Dx = A + B1lnPx + B2lnM

REPORTING THE RESULTS:

Dx = A + B1lnPx + B2lnM

Coefficient -289.3213 -14.71877 340.8743

Standard Error 283.2153 38.13928 48.47384

t-statistic -1.021560 -0.305921 7.032130

prob. value 0.3410 0.7110 0.0002

R- squared: 0.984750 Adjusted R-squared: 0.980393

F-statistic: 226.0146 Prob. (F-statistic): 0.00000

CONFIDENCE INTERVALS:
INTERPRETATION THE RESULTS:
The multiple regression model you've presented, lnD = A + bP + cM, has several statistical parameters and results
that are important for interpretation. Let's go through each of these results one by one:

1. Coefficients:

- A (Intercept) = 598.121

- b (Coefficient for P) = -2.448

- c (Coefficient for M) = 9.192

These coefficients represent the estimated relationship between the natural logarithm of D (lnD) and the predictor
variables P and M.

2. Standard Errors:

- Standard Error for A = 343.957

- Standard Error for b = 6.293

- Standard Error for c = 5.624

Standard errors measure the variability in the coefficient estimates. Smaller standard errors indicate more precise
estimates.

3. t-Statistics:

- t-statistic for A = 1.739

 - t-statistic for b = -0.389

- t-statistic for c = 1.634

The t-statistics are used to test the hypothesis that the coefficients are different from zero. In this case, the t-
statistic for A is relatively large, while the t-statistic for b is close to zero, indicating that A may be statistically
significant, but b is not. The t-statistic for c is also relatively large, suggesting that c may be statistically significant.

4. p-Values:

- p-value for A = 0.126

- p-value for b = 0.709

- p-value for c = 0.146

The p-values associated with the coefficients test the null hypothesis that each coefficient is equal to zero. A low p-
value (typically less than 0.05) indicates that the coefficient is statistically significant. In this case, A and c may be
considered statistically significant, while b is not

5. R-squared and Adjusted R-squared:

- R-squared = 0.972

- Adjusted R-squared = 0.964

R-squared represents the proportion of variance in the dependent variable that is explained by the independent
variables. In this model, a very high R-squared (close to 1) suggests that a significant portion of the variance in lnD
is explained by P and M.

6. F-Statistic:

- F-statistic = 120.391

The F-statistic is used to test whether the overall model is statistically significant. A low p-value for the F-statistic
(in this case, very close to zero) indicates that the model as a whole is statistically significant.

7. Confidence Intervals:

- 95% Confidence Intervals for M: Low = 0.347, High = 0.607

- 95% Confidence Intervals for P: Low = -0.027, High = 0.013

- 95% Confidence Intervals for A: Low = 5.287, High = 7.748

The confidence intervals provide a range within which we can be reasonably confident that the true values of the
coefficients lie. In this case, the confidence intervals for M, P, and A do not include zero, which supports their
statistical significance.

In summary, this multiple regression model suggests that variables A and M are statistically significant predictors
of lnD, while variable P is not statistically significant. The model is highly significant, and it explains a substantial
portion of the variance in lnD, as indicated by the high R-squared value. The confidence intervals for the
coefficients provide a range of plausible values for the true coefficients.