how do i find... (math, statistics ques)

[™]

the z-critical value appriopriate for a 97% confidence interval?

[WildBobAA]

You have to look at a normal distribution table.

http://people.hofstra.edu/faculty/St...rmaltable.html

You'd add .5 to the values in the table to get the confidence level.

Example: The z critical value .10 would have a 0.5398 confidence level.

[™]

my book has z critical values for confidence intervals...
90% is 1.645
95% is 1.96
98% is 2.33
99% is 2.58
99.8% is 3.09
99.9% is 3.29

I need 97%...

I dont really understand what your saying bob, it doesnt seem to work...

[gengar]

Use the link that WildBob provided, and use the chart backwards.

You want 97%. The graph starts at the middle (50%) and counts up from there, so for 97% you need 47%. So find the lowest value for 47%... there is a value for .4699, which corresponds to 1.88... so you know that 1.882 or so should do it. And there you go.

[Demiparadigm]

http://www.flopturnriver.com/phpBB2/...=188609#188609

not exactly what you're looking for...
what are you looking for?

EDIT: I think I figured out what you wanted...

[midas06]

Half of 0.97 = 0.485.

Using the CI table, z=2.17

[WildBobAA]

Half of 0.97 = 0.485.

Using the CI table, z=2.17

No it's not half you add .5 to the values.




Gengar is right. All the area less than zero corresponds to half the area under the normal curve. And the table I linked gives the area from zero to whatever z value you use it for. If you're having a problem reading the table, the z-values are listed on the very left/top of the table. The the ones/tenths values are listed on the left and the hundreths values are listed on the top. So if you wanted to look up the confidence interval for 1.12, you'd look on the left side for the row labled 1.1 and then go right to the column labled 0.02 and add .5 to the value there and that would give you the confidence interval.

So since you're looking for the .97 confidence interval look on the graph for .47 (.47 + .5 = .97). Which (like gengar said) is about 1.88.

[™]

Half of 0.97 = 0.485.

Using the CI table, z=2.17

No it's not half you add .5 to the values.




Gengar is right. All the area less than zero corresponds to half the area under the normal curve. And the table I linked gives the area from zero to whatever z value you use it for. If you're having a problem reading the table, the z-values are listed on the very left/top of the table. The the ones/tenths values are listed on the left and the hundreths values are listed on the top. So if you wanted to look up the confidence interval for 1.12, you'd look on the left side for the row labled 1.1 and then go right to the column labled 0.02 and add .5 to the value there and that would give you the confidence interval.

So since you're looking for the .97 confidence interval look on the graph for .47 (.47 + .5 = .97). Which (like gengar said) is about 1.88.

how can it be 1.88 when 95% is 1.96... i mean, the value must be between 1.96 and 2.33...

[WildBobAA]

Oh ok I know now. You're looking for the "middle" .97 in area not the left .97 in area.

Ok so then to use the chart you'd take (1 - .97) = .03. .03/2 = .015. then take .5 - .015 = .485 and find that on the chart.

So it is 2.17. midas was correct, I apologize.