Specification Tests: Part 1 (Video 4 of 7 in tbhe gretl Instructional Video Series)

The gretl Instructional Video Series consists of seven videos that instruct and demonstrate how to use gretl to apply econometric techniques. The videos are designed to be 'hands on' and will be the most effective if the students follow and actively participate using gretl on their own computer while watching the video. The videos can be used for individual study or for in-class presentation.
The learning objectives for Video 4 include the following: perform JB test for normality, Ramsey's RESET test, Chow test, White's test.
The data files used in this video can be downloaded from www.janellemann.com/uploads/8/...
Citation: Mann, J. (2011, September 1). Specification Tests: Part 1 - Video 4 of 7 in the gretl Instructional Video Series. Retrieved from • Specification Tests: P...
It is important note that the videos are not designed to teach econometric theory and techniques, rather they apply the econometric theory and techniques learned in class. Additional information on all topics covered in the videos can be found in the course lecture notes.

[@ChrisloveCGO]

These are very helpful and really help my use of gretl.Thank you!

[@Emiliano1115]

Great!

[@drgomez68]

Great instructional Video Series...Good Job

[@Kyyp3r]

very nice

[@ikramaitmansour3631]

please what should we do when we have a mispecification error in the Ramsey reset test ?

[@JanelleMann]

Hello Ikram - Two common culprits are misspecified functional form or omitted variables.

[@JanelleMann]

The best thing is to go back to theory - think about potential variables missing from the model. Also, think about the functional form.

[@femiolumide6857]

For the JB test, the null hypothesis is that the residuals are normally distributed. Therefore to accept the null the p-value must be greater that 0.05. In the video, the p-values are high i.e. greater than 0.05- therefore normally distributed. Please check your explanations for the other tests too. These are the other way round and misleading.

[@JanelleMann]

Thank you for your comment. The p-values are large, so we fail to reject the null hypothesis that the residuals are Normally distributed. Accepting and failing to reject are similar, but not exactly the same. Here is a good explanation: liesandstats.wordpress.com/2008/09/08/accept-the-null-hypothesis-or-fail-to-reject-it/

[@bartonpaullevenson3427]

The Chow test should properly be used with a wide variety of possible break points, not just the one you expect. Otherwise you bias the test.

[@Yasha-uh9vj]

Thanks for the information, do you know any other good videos like this one we learn from?

[@bartonpaullevenson3427]

@@Yasha-uh9vj Not offhand. I have a tutorial on statistics on the web, but you may find it a bit elementary: bartonlevenson.com/ISK/Statistics/00Stats.html