Correlation analysis is used to determine the degree of relationship between variables. In correlation it is assumed that the variables mutually influence each other (Sharma, 2005). Baker and Saltes (2005) used a correlation to test capability of ABI for forecasting CS sectors and then used regression analysis’s coefficient of determination (R2) to indicate the proportion of the variance of CS that is explained by ABI. They found that non-residential CS is highly correlated with ABI in lag 5 for overall sector and 11 for commercial/industrial sector. they concluded that ABI can be used for forecasting CS as a leading indicator. They pointed out that formal structural models are needed to improve the accuracy of CS forecast.
Baker, Chu, & Riskus (2014) introduced “Design Contracts” as a new leading indicator to help forecating CS. The design contracts includes information about changes in inquiries for new works including bids and requests for proposals. AIA collects data on design contracts form October 2010. They concluded that this indicator provides earlier insignt into future dedign and construction activities.
There are some studies regarding forecasting of CS. None of these studies did use explanatory …show more content…
The null hypothesis of this test is that the past p values of X are not helpful to predict Y (X does not Granger cause Y). The alternative hypothesis is that the past p values of X are useful to predict Y (X Granger causes Y). p is the lag length of the Granger causality test and the results of the test depend on the chosen lag lengths (p). Therefore, different lag lengths (6, 12, 18, 24, 30, 36 lag lengths) are used in this study. These lag lengths represent a three-year horizon. The following bivariate autoregressive model is used to test whether the ABI Granger causes
Baker, Chu, & Riskus (2014) introduced “Design Contracts” as a new leading indicator to help forecating CS. The design contracts includes information about changes in inquiries for new works including bids and requests for proposals. AIA collects data on design contracts form October 2010. They concluded that this indicator provides earlier insignt into future dedign and construction activities.
There are some studies regarding forecasting of CS. None of these studies did use explanatory …show more content…
The null hypothesis of this test is that the past p values of X are not helpful to predict Y (X does not Granger cause Y). The alternative hypothesis is that the past p values of X are useful to predict Y (X Granger causes Y). p is the lag length of the Granger causality test and the results of the test depend on the chosen lag lengths (p). Therefore, different lag lengths (6, 12, 18, 24, 30, 36 lag lengths) are used in this study. These lag lengths represent a three-year horizon. The following bivariate autoregressive model is used to test whether the ABI Granger causes