WORK 3.

20 CHAPTER 3

From the graph we can see that the trend of the two variables is increasing.

In the graph of the non-agricultural business sector we can also see that the variables have an
increasing trend.
By analyzing the relationship between the real hourly remuneration variable and the hourly
production of all people in the business sector, we can verify a positive or direct relationship
between these two variables. That is, as production per hour increases, real remuneration per
hour also increases.

When analyzing the variables real remuneration per hour and production per hour of all people for
the non-agricultural business sector we see a positive or direct relationship between these
variables. This indicates that as the hourly production of all people increases, the real hourly wage
or remuneration in the non-agricultural business sector will also increase.
Dependent Variable: SAL_HORA_SECT_BUSINESS
Method: Least Squares
Date: 04/21/18 Time: 16:22
Sample: 1960 2005
Included observations: 46

Variable Coefficient Std. Mistake t-Statistic Prob.

c -102.3662 4.503487 -22.73042 0.0000

PROD_HOUR_SECT_BUSINESS 1.992435 0.050625 39.35690 0.0000

R-squared 0.972379 Mean dependent var 69.14783

Adjusted R-squared 0.971751 S.D. dependent var 45.83186
SE of regression 7.703189 Akaike info criterion 6.963651
Sum squared resid 2610.921 Schwarz criterion 7.043157
Log likelihood -158.1640 Hannan-Quinn criter. 6.993434
F-statistic 1548.966 Durbin-Watson stat 0.123870
Prob(F-statistic) 0.000000

Performing linear regression using the ordinary least squares method, the model shows us that,
although there is a very high linear relationship, that the estimators are statistically significant and
the model shows high significance, we have to see that there is autocorrelation in the model,
which indicates that the estimators are not close to reality. However, this model corroborates the
positive or direct relationship between the variables hourly remuneration and hourly productivity
of people. The model indicates that, if the hourly productivity of all people in the business sector
increases by 1 unit of measurement, the real hourly compensation of people will increase by
approximately 2 dollars. The autonomous value is not logical at all because its interpretation does
not reflect the logic of reality since it indicates that if the productivity per hour of people is zero or
zero 0, the real salary per hour will be -102.36 dollars, which does not make any sense, since no
one works to lose, and if you do not work what would happen is that you would not earn the
salary for the day or the hour worked, it is not that you would owe the company for not working .
Dependent Variable: SAL_HORA_SECT_NEG_NO_AGRIC
Method: Least Squares
Date: 04/21/18 Time: 18:22
Sample: 1960 2005
Included observations: 46

Variable Coefficient Std. Mistake t-Statistic Prob.

c -111.6407 4.866239 -22.94189 0.0000

PROD_HOUR_SECT_NEG_NO_AGRIC 2.075734 0.054257 38.25712 0.0000

R-squared 0.970815 Mean dependent var 69.17826

Adjusted R-squared 0.970151 S.D. dependent var 45.46502
SE of regression 7.854881 Akaike info criterion 7.002652
Sum squared resid 2714.763 Schwarz criterion 7.082158
Log likelihood -159.0610 Hannan-Quinn criter. 7.032435
F-statistic 1463.607 Durbin-Watson stat 0.131917
Prob(F-statistic) 0.000000

When reviewing the regression analysis that relates the variables real remuneration per hour and
production per hour of all people, the variables correspond to the non-agricultural business sector,
we see that although it shows estimators with high statistical significance, coefficient of
determination with a high degree of goodness of fit, however, based on the statistician Durbin
Watson we can say that there is a problem of autocorrelation in the present model, so these
estimators would not be faithful representatives of reality. Even though the model is not close to
reality, it confirms the type of relationship that these two variables have, which we inferred in the
second graph above, where it shows us that these two variables have a positive or direct linear
relationship. To be clearer, we will interpret the estimators of the model which indicates that an
increase of 1 unit of measurement in the variable production per hour of all people in the non-
agricultural business sector would increase the real remuneration per hour by 2 dollars. The
autonomous estimate of this model does not have an interpretation with economic logic according
to the variables used to resolve this model, since it indicates that, if the hourly production of all
people in the non-agricultural business sector were zero or zero 0, The actual hourly
compensation would be -111.64 dollars. Which, of course, has no logic, an aspect that can already
be found in the previous model of the business sector.

These problems of interpretation of the autonomous estimator in the two previous models may be
a consequence of the autocorrelation problem detected previously, which will be solved so that
the model is a faithful reflection of reality.
WORK 3.22.

We have three variables to study, which are: the Consumer Price Index (CPI), Gold Price (Gold
Price) and the NEW YORK Stock Exchange Index (NYSE). We will study these three variables to later
define the impact on the other two variables of a variation in inflation or the consumer price
index. We cannot forget that the consumer price index is a variable that affects an entire country
and has multiplier effects and impacts on different macro and microeconomic variables.

As we see in the graphs, the consumer price index has a growing trend, the price of gold has
different variations, a lot of volatility, the New York Stock Exchange index has a growing trend with
a slight decrease in 2003 and 2004.
The scatter plot indicates a direct relationship between the consumer price index and the New
York Stock Exchange index, which means that as the consumer price index increases, the New York
Stock Exchange index also increases. York. In the case of the relationship between the consumer
price index and the price of gold, this graph indicates that there is also a direct relationship
between these variables.

Dependent Variable: GOLD_PRICE

Method: Least Squares
Date: 04/22/18 Time: 19:31
Sample: 1974 2006
Included observations: 33

Variable Coefficient Std. Mistake t-Statistic Prob.

c 215.2856 54.46850 3.952479 0.0004

CPI 1.038430 0.403782 2.571758 0.0151

R-squared 0.175837 Mean dependent var 347.5470

Adjusted R-squared 0.149252 S.D. dependent var 111.7487
SE of regression 103.0726 Akaike info criterion 12.16744
Sum squared resid 329342.7 Schwarz criterion 12.25813
Log likelihood -198.7627 Hannan-Quinn criter. 12.19795
F-statistic 6.613938 Durbin-Watson stat 0.588871
Prob(F-statistic) 0.015135

When performing the regression analysis between the variables price of gold and the consumer
price index, we verified that, although there are estimators with high statistical significance, the
coefficient of determination indicates that there is no statistical relationship between the
variables, in addition the Durbin coefficient Watson indicates that there is autocorrelation in the
model.
Dependent Variable: NYSE
Method: Least Squares
Date: 04/22/18 Time: 19:57
Sample: 1974 2006
Included observations: 33

Variable Coefficient Std. Mistake t-Statistic Prob.

c -3444.992 533.9663 -6.451703 0.0000

CPI 50.29719 3.958363 12.70656 0.0000

R-squared 0.838925 Mean dependent var 2961.194

Adjusted R-squared 0.833729 S.D. dependent var 2478.011
SE of regression 1010.442 Akaike info criterion 16.73286
Sum squared resid 31650810 Schwarz criterion 16.82355
Log likelihood -274.0921 Hannan-Quinn criter. 16.76337
F-statistic 161.4568 Durbin-Watson stat 0.189358
Prob(F-statistic) 0.000000

When performing the regression analysis by the ordinary least squares method for the index
variables of the New York Stock Exchange and the consumer price index, we noted that the
estimators are statistically significant, there is a direct positive relationship between the variables,
which We had already observed in the scatter graph, however, the model cannot be taken as close
to reality because it presents the problem of autocorrelation. But beyond the autocorrelation
problem, we note that the coefficient of determination shows that the variability of the consumer
price index explains 83.89% of the variability of the New York Stock Exchange index.

In conclusion, the variability of the price of gold does not depend on the consumer price index,
however, the New York Stock Exchange index does depend on the consumer price index. To give
you a better view of the investment panorama, we can say that as inflation increases, the New
York Stock Exchange index also increases. In the case of the price of gold, it reflects a correlation
with the consumer price index. Therefore, its variability in price is due to other variables.
WORK 3.23.

When reviewing nominal GDP and real GDP we see that both have an increasing trend.
Dependent Variable: NGDP
Method: Least Squares
Date: 04/23/18 Time: 00:30
Sample: 1959 2005
Included observations: 47

Variable Coefficient Std. Mistake t-Statistic Prob.

c -496003.3 21078.01 -23.53179 0.0000

YEAR 0.691515 0.029131 23.73819 0.0000

R-squared 0.926048 Mean dependent var 4338.632

Adjusted R-squared 0.924404 S.D. dependent var 3598.775
SE of regression 989.4704 Akaike info criterion 16.67384
Sum squared resid 44057325 Schwarz criterion 16.75257
Log likelihood -389.8352 Hannan-Quinn criter. 16.70346
F-statistic 563.5015 Durbin-Watson stat 0.034611
Prob(F-statistic) 0.000000

GDP grows as time goes by. The interpretation of this model is simple: over time for one year,
nominal GDP will grow by 0.69 trillion dollars. Although the estimators are statistically significant,
the model is also significant, the coefficient of determination indicates that over time measured in
one year it explains 92.60% of the variation in nominal GDP, however, the model has the problem
of autocorrelation. , so the estimators may be far from reality.

Dependent Variable: GDPR

Method: Least Squares
Date: 04/23/18 Time: 00:34
Sample: 1959 2005
Included observations: 47

Variable Coefficient Std. Mistake t-Statistic Prob.

c -351146.2 9065.301 -38.73519 0.0000

YEAR 0.493533 0.012529 39.39210 0.0000

R-squared 0.971818 Mean dependent var 5946.738

Adjusted R-squared 0.971191 S.D. dependent var 2507.225
SE of regression 425.5548 Akaike info criterion 14.98629
Sum squared resid 8149359. Schwarz criterion 15.06502
Log likelihood -350.1777 Hannan-Quinn criter. 15.01591
F-statistic 1551.738 Durbin-Watson stat 0.082125
Prob(F-statistic) 0.000000

The regression analysis of Real GDP indicates that as time passes by one year, real GDP increases
by 0.49 trillion dollars, the coefficient of determination indicates that the passage of time
measured in years explains 97.19% of the variation in real GDP. Although the estimators have high
statistical significance, the model suffers from the autocorrelation problem so the estimators may
be far from reality.
JOB 3.25

When reviewing the scatter graph of the scores for the SAT exam in critical reading for men and
critical reading for women we see that there could be a linear relationship between these two
variables, however, we cannot predict if there is causality between these two variables. because
there is no logic in which, if men get higher grades in this subject, women would have to increase
their grades. The only explanation of causality would be the competition between men and
women to get better grades in the subject of critical reading.
Dependent Variable: CRITICAL_READING_FEMALE
Method: Least Squares
Date: 05/01/18 Time: 17:59
Sample: 1972 2007
Included observations: 36

Variable Coefficient Std. Mistake t-Statistic Prob.

c -31.76318 47.80405 -0.664445 0.5109

CRITICAL_READING_MALE 1.048498 0.093722 11.18733 0.0000

R-squared 0.786374 Mean dependent var 503.0000

Adjusted R-squared 0.780090 S.D. dependent var 7.147427
SE of regression 3.351751 Akaike info criterion 5.310796
Sum squared resid 381.9640 Schwarz criterion 5.398769
Log likelihood -93.59432 Hannan-Quinn criter. 5.341501
F-statistic 125.1564 Durbin-Watson stat 0.526827
Prob(F-statistic) 0.000000

The analysis between women's grades versus men's grades in the subject of critical reading
indicates that there may be a relationship between women's grades and women's grades, perhaps
due to greater competition between men and women to obtain the best grades in the exams and
excel in this subject, however, there could be other causes for different relationships to exist.
WORK 3.26

When reviewing the graphs of mathematics grades for men and women, we see that, although
men stand out more in this subject, both men and women have very similar cycles over time with
respect to their grades. As for the scatter graph, we see that there is a very marked linear
relationship, but that does not mean that we can determine that there is causality, since there is
no logical reason why one variable is explained by the other. We will look at linear regression to
confirm this hypothesis.
Dependent Variable: MATHEMATICS_FEMALE
Method: Least Squares
Date: 05/01/18 Time: 21:34
Sample: 1972 2007
Included observations: 36

Variable Coefficient Std. Mistake t-Statistic Prob.

c -257.0235 29.34764 -8.757893 0.0000

MATHEMATICS_MALE 1.416420 0.055913 25.33247 0.0000

R-squared 0.949684 Mean dependent var 486.3611

Adjusted R-squared 0.948204 S.D. dependent var 10.10277
SE of regression 2.299251 Akaike info criterion 4.556997
Sum squared resid 179.7430 Schwarz criterion 4.644970
Log likelihood -80.02595 Hannan-Quinn criter. 4.587702
F-statistic 641.7338 Durbin-Watson stat 1.183041
Prob(F-statistic) 0.000000

The linear regression of the model by the ordinary least squares method shows us that there is a
very high linear relationship between the mathematics grades of men and women, the coefficient
of determination indicates that the variations in the mathematics grades of men explain in a
94.96% to variations in women's mathematics grades. This rejects our hypothesis that there would
be no causality between these two variables, however in terms of the subject of mathematics
there is competition that especially women highlight for wanting to be in equal opportunities to
have jobs in the technological sector.