My Account Order Now
x

Request Call Back

writing-offer
FREE INSTANT QUOTEClick Here

Great 9.5/10

Trust PilotIgnite ReviewSite JabberReview Centre
Cross

LIVE CHAT

How can we help?

How to take a Unit Root Test?

dissertation-software

This blog has been specifically developed in the continuation of the last blog posted about which quantitative test and software to pick for a particular dissertation subject or topic. To further continue providing dissertation help, this blog mainly addresses how to take a unit root test if you intend to employ it in your economics dissertation.  Being a student of Master’s or PhD levels, your assessors and tutor anticipate it from you to employ a high class quantitative test in your analysis to process the data. Particularly, if you’re student of Economics or Econometrics, you must have been asked to carry out a number of quantitative tests, including Regression model, Correlation, ANOVA, Multivariate test.

Unit Root test is one of the tests chiefly adopted by students to conduct on the variables such as Inflation, Corruption Perceived Index,% of Gross Domestic Product (GDP), FDI inflow, political instability, natural resources, and so on. Time series analysis is the significant aspect of the unit root test. The data is always required to be stationary when it comes to time series analysis. The role of the unit root test in these remains the determiner of the stationary and non-stationary of the series being concerned. Going further, a question that arises is“Why is this test for non- stationary need?”the answer to which is very simple, that the stationary of the series being assume has a definite impact on the properties and the behavior of the series. For instance, the determination of the shocks in the non-stationary series remains infinite, but in stationary have a proper start and an appropriate end.

What should I start with?

The most initial step for the unit root test remains consideration of the trend cycle decomposition of the time series.

yt= TDt + TSt + Ct

yt= TDt + Zt

This equation underlies the idea that the unit root test determines whether the T St= 0. For the resolution of this, there are two hypotheses constructed for testing the stationary and non- stationary of the series, named as the unit root test. 

H0: T St 6= 0 (yt∼ I (1)) vs. T St= 0 (yt∼ I (0))

Or

H0: T St= 0 (yt∼ I (0)) vs. T St 6= 0 (yt∼ I (1))

What else could be done?

For the test of unit root the Autoregressive Unit Root Tests are being emphasized, which are based on the idea that: 

yt=ø yt−1 + ut, ut∼ I (0)

The testing here is laid on the hypothesis that:

H0:ø= 1 (which means,ø (z)= 0, it has a unit root)

H1:|ø|< 1 (which means,ø (z)= 0, the roots are outside unit circle)

What is the best procedure to be applied?

For the accomplishment of the unit root test, there are two eminent procedures or tests that could be applied to the required unit root values. They are Dickey-Fuller (ADF) test and Phillips-Peron (PP) test. Here the ADF is parametric autoregressive but PP tests are non-parametric autoregressive structures.

What is the actual test?

The above determined tests are the core tests for the unit root test, while their interpretation is based on two corresponding ways, which are World representation and UC-ARIMA model. Application of these in the software is also deliberately found provided producing appropriate results.

Which Software helps in this test?

There are various statistical software  that are found providing with the testing of the unit root tests, in which the eViews, STATA, R, Excel and many others. They just require an appropriate data to be fed into the software and then through simple direct, methods any of the unit root tests could be performed in no time. The important point to be noted is that SPSS (our most favourite and easiest statistical software) doesn’t support the function required for carrying out the Unit Root Test.

What data is required for this test?

It is quite coherent that for this testing process data would be required, for the unit root testing the data could be any financial or economic indicator with specification of having their trends defined. This is important because the data with no trends is obvious to include absence of time series would exclude it from the testing of Unit Root Test. Best example of data could be inflation rate of a United Kingdom from 1980 to 2010. Another example of the data is Corruption Perceived Index of South Africa from 1996 to 2016.

We hope that you enjoy and learn more about unit root test. You will enjoy it more when you apply it to your dissertation. If you are still lacking the fundamental understanding of the test, you can ask our experts to help you more. You can also buy assignment service from our expert.

10/3/2017 6:07:53 PM
Review posted and now it's pending for approval!Review posted and now it's pending for approval!

Page Reviews