###
**CFA Practice Question**

When trying to increase the power of a test, the probability of ______

A. Type I error is automatically reduced.

B. rejecting the null when it is true increases.

C. Type II error is increased.

**Explanation:**Power of a test = 1 - Prob(Type II error). The power is maximized when the probability of Type II error is minimized. This means that the alternate hypothesis is rejected less often when it is true and the null hypothesis is rejected more often when it is true.

###
**User Contributed Comments**
14

User |
Comment |
---|---|

danlan |
Power of test is 95% at significance of 95%. Think of this. |

danlan |
proba of type I error increased |

danlan |
Power of test is 95% when significance level of 1-95% when trying to increase the power of a test, proba of type II error reduced and proba of type I error increased. when trying to increase the significance level, proba of type II error increased and proba of type I error decreased. |

srinib |
power of test is 1 - alpha? |

achu |
Choices 1 and 3 go together so they can't be right. By definition of POWER of a test choice B is clearly the right one. |

dealsoutlook |
danlan you dont make sense. You said that when trying to increase significance level probability of type 1 error is decreased, how is that possible? Isnt significance level represented by alpha which is type 1 error? which would mean as you increase significance level type 1 error would increase. I am really confused in stats can someone PLEASE clarify |

steved333 |
deals, you're right. danlan's apparently a pretty smart guy, but so am I. I'm wrong about many things. danlan's wrong about this one. Power of test is the significance level, which is 1-confidence level. Easy to confuse the two... |

StanleyMo |
for power of test, we have 1 - beta, where beta is the confidence level, to improve the power of test, we reduce beta, where reduce area of confidence area, and the significance level increase, and increase the probability of rejecting the null ( either is true or not true). |

rana1970 |
there is trade off between type-1 and type-11 error. to reduse beta, u can increase alpha, thus increase power of test. so B is riht choice |

mattg |
Type III error: staring at Analystnotes until you go cross-eyed and crazy working on Quantitative mumbo-jumbo! |

soukhov |
good one, mattg =)) |

Clude |
Think this way,extreme case is thay P(type 2)=0, which means "impossible to accept null when its false". Then "imposisble to accept" = "always reject". |

cslau83 |
power of test = is the probability of correctly rejecting the null. If power of test increase isn't the probability of Type I error reduced?? Meaning the answer is A? |

Rohule |
more false positives |