7 thoughts on “13.4 | F-test & Estimate (Preview)”

Jpiercy10

What do you do when looking up v1=29 & v2 =29 for an F score. Do you look at 30 (the higher value) or do you take 28 & 30 and average the two? Came across this situation in question 13.76 in the textbook

It is safer to choose the lower degrees of freedom when the specific degrees of freedom you’re looking for are not listed on the chart.

Why? There are only a few options available to you:

Take the average of the values for 28 and 30, and then average them.

Use the higher value 30 which represents a larger sample.

Use the lower value 28 which represents a smaller sample.

Option 1 is not a bad idea, but the method is time consuming, and does not produce as accurate of a result as you’d hope for (the value for 29 is not simply the average of the values for 28 and 30).

Option 2 requires that you use a value that would only have occurred naturally if you had a sample that was larger than your current sample. This puts more faith in your sample than it actually deserves and is considered a bad practice.

Option 3 is more conservative than option 2, and is faster than option 1, so it’s the right choice.

A menâ€™s softball league is experimenting with a yellow baseball that is easier to see during night games. One way to judge the effectiveness is to count the number of errors. In a preliminary experiment, the yellow baseball was used in 10 games and the traditional white baseball was used in another 10 games. The number of errors in each game was recorded and is listed here. Can we infer that there are fewer errors on average when the yellow ball is used?
Yellow 5 2 6 7 2 5 3 8 4 9
White 7 6 8 5 9 11 8 3 6 10

Possible if you can please explain how to get the solution? Solution Manual does a terrible job

Hi Jason,
I am a little confused from the f-test and the f- estimate. For the rejection region, you had said to look up f, subscript alpha, v1, v2. but for the ucl and lcl, you have to look up f subscript alpha divided by 2, v1, v2. Is that correct? in other words are they supposed to be different?

What do you do when looking up v1=29 & v2 =29 for an F score. Do you look at 30 (the higher value) or do you take 28 & 30 and average the two? Came across this situation in question 13.76 in the textbook

It is safer to choose the lower degrees of freedom when the specific degrees of freedom you’re looking for are not listed on the chart.

Why? There are only a few options available to you:

Option 1 is not a bad idea, but the method is time consuming, and does not produce as accurate of a result as you’d hope for (the value for 29 is not simply the average of the values for 28 and 30).

Option 2 requires that you use a value that would only have occurred naturally if you had a sample that was larger than your current sample. This puts more faith in your sample than it actually deserves and is considered a bad practice.

Option 3 is more conservative than option 2, and is faster than option 1, so it’s the right choice.

A menâ€™s softball league is experimenting with a yellow baseball that is easier to see during night games. One way to judge the effectiveness is to count the number of errors. In a preliminary experiment, the yellow baseball was used in 10 games and the traditional white baseball was used in another 10 games. The number of errors in each game was recorded and is listed here. Can we infer that there are fewer errors on average when the yellow ball is used?

Yellow 5 2 6 7 2 5 3 8 4 9

White 7 6 8 5 9 11 8 3 6 10

Possible if you can please explain how to get the solution? Solution Manual does a terrible job

Thanks

I put together a walk-through of how the f-test part of this solution works.

Hi Jason,

I am a little confused from the f-test and the f- estimate. For the rejection region, you had said to look up f, subscript alpha, v1, v2. but for the ucl and lcl, you have to look up f subscript alpha divided by 2, v1, v2. Is that correct? in other words are they supposed to be different?

Yes – They must be different. Check this out:

That was extremely helpful, thank you very much for the quick response