In: Statistics and Probability
A.) Suppose that we run the exact same experiment 100 times and find a statistically significant difference 30 times out of 100 using an alpha level of .05. What is the best estimate of our statistical power?
a) a difference between means is very likely to be detected
b) the significance level set by the researcher must be high
c) the probability of type I error is very high
d) Your study will likely be inconclusive
B) A researcher believes that dogs can smell fear, so he conducts an experiment with blindfolded dogs next to a person who is watching either a scary or a funny movie. What is the null hypothesis (Ho)?
a) dogs can smell fear
b) dogs cannot smell fear
c) that the probability value will be below 0.05
d) that there will be a statistically significant result
A.) Suppose that we run the exact same experiment 100 times and find a statistically significant difference 30 times out of 100 using an alpha level of .05. What is the best estimate of our statistical power?
So 30 time out of 100 we find statistically significant difference.
Statistical power, or the power of a hypothesis test is the probability that the test correctly rejects the null hypothesis. i.e the probability of accepting the alternative hypothesis if it is true.
So we have rejected null hypothesis 30 time out of 100 times.
Now Power = P ( Reject H0 | H1 is true )
Since using an alpha level of .05 , runing the same experiment , we find a statistically significant difference 30 times out of 100 ( which implies reject H0 30 times ) whereas it should be 5 times or less out of 100.
Hence correct option is
a) a difference between means is very likely to be detected
B) A researcher believes that dogs can smell fear, so he conducts an experiment with blindfolded dogs next to a person who is watching either a scary or a funny movie. What is the null hypothesis (Ho)?
Here null and alternative hypothesis will be
H0 : dogs can smell fear
H1 : dogs can not smell fear
Hence correct option is
a) dogs can smell fear