#PROVIDING THE DESCRIPTIVE STATICTICS

summary(total_cost_labor)
V1
Min. : 500
1st Qu.: 30000
Median : 84920
Mean : 1104954
3rd Qu.: 455600
Max. :100000000

> summary(labor)
V1
Min. : 1.00
1st Qu.: 8.00
Median : 15.00
Mean : 38.58
3rd Qu.: 35.00
Max. :1000.00

# VISUALISING THE VARIABLES

# LINEAR REGRESSION ANALYSIS

Call:
lm(formula = total_cost_labor ~ labor, data = CLASS_TEST_ONE_DATASET)

Residuals:
Min 1Q Median 3Q Max
-25229428 -415268 105329 305329 63619952

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -499174 207740 -2.403 0.0165 *
labor 41584 2414 17.228 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 4716000 on 643 degrees of freedom

Multiple R-squared: 0.3158, Adjusted R-squared: 0.3147
F-statistic: 296.8 on 1 and 643 DF, p-value: < 2.2e-16

> coef(lm_model)
(Intercept) labor
-499173.57 41584.43

#NON-LINEAR REGRESSION ANALYSIS (QUADRATIC)

Call:
lm(formula = total_cost_labor ~ I(labor^2), data = CLASS_TEST_ONE_DATASET)

Residuals:
Min 1Q Median 3Q Max
-30515340 -522283 -480305 -237081 64476396

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 5.293e+05 1.696e+05 3.122 0.00188 **
I(labor^2) 7.772e+01 3.441e+00 22.588 < 2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 4258000 on 643 degrees of freedom

Multiple R-squared: 0.4424, Adjusted R-squared: 0.4416
F-statistic: 510.2 on 1 and 643 DF, p-value: < 2.2e-16

> coef(nls_model)
(Intercept) I(labor^2)
529309.1079 77.7178

# LOG-LOG REGRESSION
Call:
lm(formula = log(total_cost_labor) ~ log(labor), data =
CLASS_TEST_ONE_DATASET)

Residuals:
Min 1Q Median 3Q Max
-5.0469 -1.0155 0.1027 0.9281 5.1964
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 8.46437 0.16963 49.90 <2e-16 ***
log(labor) 1.10740 0.05469 20.25 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 1.523 on 643 degrees of freedom

Multiple R-squared: 0.3893, Adjusted R-squared: 0.3884
F-statistic: 410 on 1 and 643 DF, p-value: < 2.2e-16

> coef(log_model)
(Intercept) log(labor)
8.464371 1.107399

#REGRESSION RESULTS BY USING THE OUTREG CODE

.variable .stat Model 1 Model 2 Model 3

1 (Intercept) Estimate -499173.573** 529309.108*** 8.464***
2 Std Err [207740.072] [169565.941] [0.170]
5 labor Estimate 41584.433***
6 Std Err [2413.811]
9 I(labor^2) Estimate 77.718***
10 Std Err [3.441]
13 log(labor) Estimate 1.107***
14 Std Err [0.055]
17 N 645 645 645
18 R2 0.316 0.442 0.389
19 adj R2 0.315 0.442 0.388
20 AIC 21657.255 21525.245 2377.141

INTERPRETATION

Model 1:

 Intercept ( (Intercept) Estimate ): The intercept is -499173.573. This suggests that

when all predictor variables (labor, labor squared, log(labor), etc.) are zero, the estimated
value of the dependent variable (which isn't explicitly stated here) is approximately -
499,174.
 labor ( labor Estimate ): The coefficient for labor is 41584.433. This means that for
every unit increase in the labor variable, holding other variables constant, the dependent
variable is estimated to increase by 41,584.433.
 I(labor^2) ( I(labor^2) Estimate ): This coefficient is not provided in Model 1,
indicating that the quadratic term of labor is not included in this model.
 log(labor) ( log(labor) Estimate ): This coefficient is also not provided in Model 1,
indicating that the natural logarithm of the labor variable is not included in this model.

Model 2:
  Intercept ( (Intercept) Estimate ): The intercept is 529309.108. This suggests that when
all predictor variables are zero, the estimated value of the dependent variable is
approximately 529,309.
 labor ( labor Estimate ): The coefficient for labor is not provided in Model 2, indicating
that the labor variable might not be included as a predictor in this model.
 I(labor^2) ( I(labor^2) Estimate ): The coefficient for the quadratic term of labor is
77.718. This means that the relationship between the dependent variable and labor is not
linear; it's quadratic, suggesting a curve in the relationship.
 log(labor) ( log(labor) Estimate ): This coefficient is not provided in Model 2,
indicating that the natural logarithm of the labor variable is not included in this model.

Model 3:

 Intercept ( (Intercept) Estimate ): The intercept is 8.464. When all predictor variables
are zero, the estimated value of the dependent variable is 8.464.
 labor ( labor Estimate ): This coefficient is not provided in Model 3, indicating that the
labor variable might not be included as a predictor in this model.
 I(labor^2) ( I(labor^2) Estimate ): This coefficient is not provided in Model 3,
indicating that the quadratic term of labor is not included in this model.
 log(labor) ( log(labor) Estimate ): The coefficient for log(labor) is 1.107. This means
that for every unit increase in the natural logarithm of the labor variable, holding other
variables constant, the dependent variable is estimated to increase by 1.107.

Summary:

 Model 1 includes an intercept and a linear relationship with the labor variable.
 Model 2 includes an intercept, a quadratic relationship with the labor variable, but no
linear relationship (labor itself is not included).
 Model 3 includes only an intercept and a logarithmic relationship with the labor variable.

These interpretations are based solely on the coefficient estimates provided in each model. They
indicate the direction and magnitude of the estimated relationships between the predictor
variables and the dependent variable (though the exact nature of the dependent variable isn't
specified in your question).

# DIAGNOSTIC PLOTS FOR TESTING THE OLS PROPERTIES

#ELASTICITY FOR THE LINEAR REGRESSION

[1] "Elasticity of total cost of labor with respect to labour: 1.4518"

Interpreting the Elasticity Value (1.4518)

 An elasticity value of 1.4518 indicates that the total cost of labor is elastic with respect to
changes in labor quantity.
 Elasticity values greater than 1 suggest that the percentage change in total cost of labor is
proportionally greater than the percentage change in labor quantity. In other words, a 1%
increase in labor quantity leads to a 1.4518% increase in total cost of labor.
 This suggests that the total cost of labor is responsive to changes in labor quantity; as more
labor is employed, the total cost increases at a rate greater than the rate of increase in labor
quantity.

Practical Implications

 Management of Labor Costs: Understanding the elasticity helps managers and policymakers
anticipate the financial implications of changes in labor quantity. For example, when planning
workforce expansions or contractions, they can predict how total labor costs will change.
  Cost-Benefit Analysis: Businesses can use elasticity to assess the cost-effectiveness of hiring
additional labor. A high elasticity might indicate that hiring more labor could significantly
increase costs, influencing decisions on workforce size.
 Policy Decisions: Policymakers can use elasticity to evaluate the economic impact of labor-
related policies, such as minimum wage changes or labor regulations, on total labor costs.

In summary, an elasticity of 1.4518 for the total cost of labor with respect to labor quantity
signifies that changes in the quantity of labor employed have a significant impact on total labor
costs, with costs increasing at a rate greater than the rate of increase in labor quantity.

# R_SQUARED CALCULATIONS

#R-squared Manual calculation

> SST <- sum((CLASS_TEST_ONE_DATASET$`Total cost of labor`- mean_Y)^2)

> SSE <- sum(residuals(lm_model)^2)
> R_square_manual <- 1- SSE/SST
> print(paste("R_square: ", round(R_square_manual, 4)))
[1] "R_square: 0.3158"

# R_SQUARED REPORTED

R-squared: 0.3158