Statistics: paried observations

[k77i]

Homework Statement

To test the efficacy of a new cholesterol-lowering medication, 10 people are selected at random. Each has their LDL levels measured (shown below as Before), then take the medicine for 10 weeks, and then has their LDL levels measured again (After).

Before After
195 175
131 129
141 105
182 184
173 138
123 120
147 115
156 144
125 87
161 137

Give a 95.8% confidence interval for meanB - meanA , the difference between LDL levels before and after taking the medication.

Homework Equations

d(bar) - t(alpha/2)(Sd/sqrt(n)) < meanDifference < d(bar) + t(alpha/2)(Sd/sqrt(n))

The Attempt at a Solution

I found all the values using excel:

d(bar) = 20
t(alpha/2) = 2.368676
Sd = 190.36
n = 10 (given)

and my answer was (-122.586 < meanDifference < 162.586)

but this is incorrect

 

[Ray Vickson]

Homework Statement

To test the efficacy of a new cholesterol-lowering medication, 10 people are selected at random. Each has their LDL levels measured (shown below as Before), then take the medicine for 10 weeks, and then has their LDL levels measured again (After).

Before After
195 175
131 129
141 105
182 184
173 138
123 120
147 115
156 144
125 87
161 137

Give a 95.8% confidence interval for meanB - meanA , the difference between LDL levels before and after taking the medication.

Homework Equations

d(bar) - t(alpha/2)(Sd/sqrt(n)) < meanDifference < d(bar) + t(alpha/2)(Sd/sqrt(n))

The Attempt at a Solution

I found all the values using excel:

d(bar) = 20
t(alpha/2) = 2.368676
Sd = 190.36
n = 10 (given)

and my answer was (-122.586 < meanDifference < 162.586)

but this is incorrect

I don't know what your Sd is, but if it is the sample standard deviation of the difference, then it is wrong. I get sample mean difference = m = 20, sample standard dev. of diff. = s = 15.3695 (your Sd?). This gives a confidence interval very different from yours.

RGV

 

[k77i]

yes it was the standard deviation of the means and I don't get why it would be wrong, I used excel to calculate it..

Anyways the correct interval was (8.5,31.5)

 

[Ray Vickson]

I first used Maple to do the calculation. Then, to see if Excel was giving you the trouble, I re-did the analysis using Excel. It agreed with the Maple calculation and disagreed with yours. I suggest you search for a typo or something similar.