There is an assumption implicit but fundamental theory behind the decline by the time series used in the estimate, namely that the time series has a sleep feature Stationary or the language of Engle Gernger chains are integrated level zero, which indicated by I (0). It is well known, for example, tables of t-statistic is designed primarily to deal with the results of the regression that uses static strings. This assumption has been previously treated as an axiom the mid-seventies, where researchers are conducting studies of applied without taking into account the properties of time series used prior to the assessment, was to accept the results of these tests Bmanueh and delivery capabilities based on the applicability of the theory of statistical inference on these capabilities.
But the scientists Swedes Granger and Newbold 1974 blew up real surprise when, where researchers generate the time series random static Stationary Non (specifically chains conduct random) using the method of simulation of this series does not reflect any variable is known and then considered these chains separately. And then they make a large number of estimates of the regression using these strings on each other.

After the estimate was calculated values ​​of statistic t and under the assumption that the parameter real equal to zero (ie, the parameter estimates from the regression likes to be insignificant to the independence and random variables used in the estimate), but in spite of the fact that the time series was random and independent, the researchers found that zero hypothesis that the true parameter equal to zero was rejected or the possibility of greater frequency than expected by the theory were accepted moral relationship statistically, the researchers also noted that the residue resulting from the regression estimates by a large positive self-link. The researchers reached this conclusion important and serious that the capacity and statistical tests that result from the regressions used the time series is still considered the results of improper or false decline spurious regressions can not be reassuring to the results of statistical inference on its resources. The form of this research the starting point for new research in the field of sleep testing strings, cast doubt on the results of all previous standard tests used did not take the time series properties of time series into account by the estimate.

As for the current study, the test was intended to stabilize some of the activities in Iraq and the commodity involved (agriculture, manufacturing, construction, electricity and water) to estimate the models of integration between the cross-sectional data and time series.

On this basis, the research contained in the theoretical side Mbgesan, in the first part, the research methodology was presented by showing the importance of research which confirmed the account on the new addition to the specialists and researchers in this field. The research problem can be summarized in a statement after stabilizing the merged data on the results of the assessment method of least squares using the two-stage compact. While the aim of the research included the assessment of model integration between the data cross-sectional and time series using the GDP variable is supported and all of (the number of workers, investment) independent variables indicate the extent to stabilize the time series on the results of the assessment in order to reach more accurate results from a statistical standpoint . In addition, ensure the topic on the assumptions and the nature of the variables used in it. In the second part, the researchers introduced the concept of testing Dickey Fuller used in detecting the presence of the root of the unit (Unit root) and applied to the data used in the research, based on a program of economic measurement (Eviews 5.1) and present the results of different cases for each test and the presence or absence (Cutter, the general trend) to the data of some commodity activities in Iraq.
The practical side has included a summary of the results of the assessment using the least squares method combined with a comparison between the two cases (non-stability,, stability,) in the presence and the absence of fixed effects of periods and groups.

Finally, the researchers offer conclusions and recommendations reached by the research.