Upcoming Webinars
Access the latest trends and insights shaping the profession—free for members.
.png)
 (1).jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.png)
.jpg)
.jpg)
.jpg)
Document
Notice of Rejection of All Bids.doc
Year Created: 2025
DOCUMENT
ART - Procurement Procedure, Competition and Final Unit Price, The Case of Commodities, 2016
- YEAR CREATED: 2016
The document discusses the analysis of a dataset related to the procurement procedure, competition, and final unit price of commodities. The document focuses on testing various assumptions and hypotheses related to the dataset.
First, the document presents the variance inflation factor (VIF) for electricity and gas, which measures the extent to which the variance of the estimated regression coefficients is increased due to multicollinearity. The mean VIF values for electricity and gas are reported as 1.33 and 1.22, respectively.
Next, the document tests the homoscedasticity of residuals, which refers to the assumption that the variance of the residuals is constant for any value of the explanatory variable. The Breusch-Pagan test is used, and it rejects the hypothesis of homoscedastic residuals for both electricity and gas.
Although heteroscedastic residuals are found, the document states that they do not cause any bias in the estimations. However, in order to assess statistical significance, robust standard errors need to be applied, and t-statistics can be used.
The normality of the residuals is also tested using the Shapiro-Wilk test. The test rejects the hypothesis of normally distributed residuals for both electricity and gas.
Despite the non-normality of residuals, the document concludes that the dataset is sufficiently large for the OLS (ordinary least squares) estimators to satisfy asymptotic normality. Therefore, the use of t- and F-statistics for hypothesis testing is deemed possible. The document also mentions that while some nonlinear unbiased estimators may have a smaller variance, the simple OLS method is sufficient for testing the hypotheses stated in the paper.
Lastly, the document discusses the correct model specification. The squares of the fitted values of the logarithm of the final unit price are tested for significance. The results show that the squares of the fitted values are insignificant, suggesting that no squares of the explanatory variables are missing and that the relationships in the model appear to be linear.