- CFA Exams
- CFA Level I Exam
- Topic 1. Quantitative Methods
- Learning Module 10. Simple Linear Regression
- Subject 1. Estimation of the Simple Linear Regression Model
CFA Practice Question
Which of the following statement(s) with respect to regression analysis are true?
II. If the relationship between two variables changes, only the regression coefficient in the equation will need to be changed.
III. Sometimes when one of the linear regression assumptions is not met, the regression results will not necessarily be deemed as invalid.
IV. A regression analysis may only be used for linear relationships.
I. Once a relationship is established between two variables, its exploitation will erode the findings of the regression.
II. If the relationship between two variables changes, only the regression coefficient in the equation will need to be changed.
III. Sometimes when one of the linear regression assumptions is not met, the regression results will not necessarily be deemed as invalid.
IV. A regression analysis may only be used for linear relationships.
A. I and III
B. III and IV
C. I, III and IV
Explanation: II is false because as the relationship between two variables changes, an entire new regression equation will need to be constructed.
IV is false because some non-linear relationships may be converted into a linear relationship by means of a mathematical transformation. So while a regression analysis assumes a linear relationship, it is possible to use regressions to analyze a non-linear relationship.
User Contributed Comments 7
| User | Comment |
|---|---|
| zwer | Picked A by elimination, but still have no clue what I implies. Anyone? |
| achu | Exponential regressions can be done using transformations- think of "%increase linear regressions done on LN y instead of y! |
| uberstyle | what is meant by exploitation? Relationship will erase as markets adjust? |
| bmeisner | Yes, exploitation has the meaning you refer to uberstyle. If some discovers a relationship and uses it to make money, that person's act of exploiting the relationship will cause it to erode over time. |
| Profache | Exploitation may cause bias, and bias erode a statistical model. |
| SMcalister | I'm not too happy about IV). The whole purpose of the natural log transformation is to make a linear relationship. I can see how this can be argued both ways. |
| chriswwu | I disagree with #1 re exploitation. It is often true but not always true. "It will erode" should be changed to "It will often erode" or even "It will usually erode". It should be phrased in a difference way - "It will not erode ..." and be related as a False selection. |