- CFA Exams
- CFA Level I Exam
- Topic 1. Quantitative Methods
- Learning Module 3. Model Misspecification
- Subject 2. Heteroskedasticity
CFA Practice Question
When we ignore heteroskedasticity, we tend to
B. find no significant relationships when there is one.
C. find no significant relationships when there is none.
A. find significant relationships when there is none.
B. find no significant relationships when there is one.
C. find no significant relationships when there is none.
Correct Answer: A
When we ignore heteroskedasticity, the t-statistics will be larger than it should be, making us believe that there is a significant relationship when none exists.
User Contributed Comments 2
User | Comment |
---|---|
nsmwaura | Type 1 error...reject the null hypothesis while its true |
quanttrader | heteroskedasticity implies more distrib amongst the error term -- ie std error increases, hence t value will increase causing us to reject the null when it may indeed be true (ie type 1 error-- falsely assuming a sign relationship) |