I have 100 boxes and on each of them there is a button. When you press the button, a random number appears on a screen on the box. In 99 of the 100 boxes, the random number is chosen from a normal distribution with expected value μ1 = 7 and standard deviation σ1 = 3. We call these boxes Model 1. In the last box, the random number is selected from a normal distribution with expected value μ2 = 8 and standard deviation σ2 = 2. We call these boxes Model 2. A random box is selected and I press the button a few times before claiming that this is a Model 1 box. You're not sure if you should trust me, and would like to inspect the box yourself. (a) Formulate a hypothesis test where the null hypothesis is that this is a Model 1 box. What kind of hypothesis test (normal distribution, t-distribution, χ2 distribution) Do you use and why? What do you choose as the significance level and why? Is Is this a one-sided or two-sided hypothesis test? Whether one-sided, right-sided or left sided? (b) If you can press the box's button 10 times, what must be the sum of values be for you to reject the null hypothesis? (c) You press the button 10 times, and get that the sum of the values is 102. Before you can reject the null hypothesis, I claim that this sum or higher is actually more likely if it is a Model 1 box than if it is a Model 2 box. Explain why this might be true. Rain looked the probabilities of getting a sum of 102 or higher for both a Model 1 box and a Model 2 box. (d) Explain what it is about this situation that makes the hypothesis test wrong to use if the sum becomes very high.
I have 100 boxes and on each of them there is a button. When you press the button, a random number appears on a screen on the box.
In 99 of the 100 boxes, the random number is chosen from a
with
Model 1. In the last box, the random number is selected from a normal distribution
with expected value μ2 = 8 and standard deviation σ2 = 2. We call these boxes
Model 2.
A random box is selected and I press the button a few times before claiming
that this is a Model 1 box. You're not sure if you should trust me, and
would like to inspect the box yourself.
(a) Formulate a hypothesis test where the null hypothesis is that this is a Model
1 box. What kind of hypothesis test (normal distribution, t-distribution, χ2 distribution)
Do you use and why? What do you choose as the significance level and why? Is
Is this a one-sided or two-sided hypothesis test? Whether one-sided, right-sided or
left sided?
(b) If you can press the box's button 10 times, what must be the sum of values
be for you to reject the null hypothesis?
(c) You press the button 10 times, and get that the sum of the values is 102.
Before you can reject the null hypothesis, I claim that this sum or
higher is actually more likely if it is a Model 1 box than if
it is a Model 2 box. Explain why this might be true. Rain looked
the probabilities of getting a sum of 102 or higher for both a Model
1 box and a Model 2 box.
(d) Explain what it is about this situation that makes the hypothesis test wrong
to use if the sum becomes very high.
Step by step
Solved in 5 steps
Thanks:) I also wonder where/how you got the number 1.96? Is it from the table with values for
At step 2 you got 1,96*(3/sqrt(10)) = 1,17. Isn't it 1,86?