Tut7 transformations
-
Upload
school-of-economics-north-west-university -
Category
Education
-
view
674 -
download
1
description
Transcript of Tut7 transformations
![Page 1: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/1.jpg)
ECON321Economic Analysis
Tutorial 7
Transformations
![Page 2: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/2.jpg)
Introduction• The lecture on functional forms explains that you
may want to transform your data and estimate non-linear relationships.
• This tutorial will show you how to do those transformations, but for more about the interpretations please refer to the chapter in Gujarati and the lecture PowerPoint slides.
• We use the model of house price (measured in dollars) as a function of:o ROOMS: The average number of rooms in houses.o NOX: The amount of nitrogen oxide in the air in parts per
million.o DIST: The distance of the community from employment
centres in miles.o STRATIO: The average student-teacher ratio of schools in
the community.
![Page 3: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/3.jpg)
![Page 4: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/4.jpg)
Using logs
• It maybe that the relationship is not this linear in parameters.o For example, as pollution gets higher and
higher, the associated decrease in the price become larger.
• You can check this by using a log-level model.
![Page 5: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/5.jpg)
![Page 6: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/6.jpg)
![Page 7: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/7.jpg)
![Page 8: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/8.jpg)
![Page 9: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/9.jpg)
![Page 10: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/10.jpg)
![Page 11: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/11.jpg)
![Page 12: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/12.jpg)
Using logs
Model Dependent variable
Independent variable
Interpretation of β1
Level-level y x ∆y = β1∆x
Level-log y log(x) ∆y = (β1/100)%∆x
Log-level log(y) x %∆y = (100*β1)∆x
Log-log log(y) log(x) %∆y = β1%∆x
![Page 13: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/13.jpg)
Using quadratics
• In addition to logs, you may want to use quadratics to examine increasing or decreasing returns.o For example, there may not ne a straight-line
positive relationship between the number of rooms and PRICE.
• You can check this by adding a quadratic term.
![Page 14: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/14.jpg)
![Page 15: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/15.jpg)
![Page 16: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/16.jpg)
![Page 17: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/17.jpg)
Using quadratics
• You are interested in the effect of ROOMS on log(PRICE):
• Note, the coefficient on ROOMS is negative, and on SQROOMS it is positive.
• This means that at low values of ROOMS an additional room has a negative effect on log(PRICE).
• At some point the effect becomes positive and the semi-elasticity of PRICE with respect to ROOMS is increasing as ROOMS increase.
• You can determine that turning point by the formula• In this case it is
xxyxxy 2121
ˆ2ˆˆ so ,ˆ2ˆˆ
)ˆ2/(ˆ21
* x
4.4)]062(.2/[545.* ROOMS
![Page 18: Tut7 transformations](https://reader033.fdocuments.us/reader033/viewer/2022061201/54796fcab4af9f9b158b483c/html5/thumbnails/18.jpg)
You now know how to estimate models and interpret the results. The rest of the course will be about tests to check whether the CLRM assumptions hold and the corrective steps that you need to take when they do not.
Save your workfile and continue…