Sample size for a given margin of error for a mean | AP Statistics | Khan Academy — Transcript
Full transcript
- 0:00nadia wants to create a confidence
- 0:02interval to estimate the mean driving
- 0:05range for her company's new electric
- 0:07vehicle
- 0:08she wants the margin of error to be no
- 0:11more than 10 kilometers at a 90 percent
- 0:14level of confidence a pilot study
- 0:17suggests that the driving ranges for
- 0:19this type of vehicle have a standard
- 0:21deviation of 15 kilometers
- 0:24which of these is the smallest
- 0:26approximate sample size required to
- 0:29obtain
- 0:30the desired margin of error so pause
- 0:33this video and see if you can think
- 0:35about this on your own
- 0:38so the traditional way that we would
- 0:39construct a margin of error on a
- 0:42confidence interval we take a sample and
- 0:44from that sample we construct
- 0:47the mean
- 0:48and then we add or subtract a margin of
- 0:51error around that to construct the
- 0:54confidence interval and the way that
- 0:55we've done that since we're dealing with
- 0:57means is we say all right if we don't
- 1:00know the standard deviation of the
- 1:01population it's appropriate to use the t
- 1:05statistic so our critical value we
- 1:07denote as t star and you'd multiply that
- 1:10times the sample standard deviation
- 1:13divided by the square root of
- 1:17your sample size
- 1:19now this question is all about what is
- 1:22an appropriate sample size given that we
- 1:25want to have a 90
- 1:26level of confidence and what's tricky
- 1:29here is when you're using a t-table
- 1:31right over here not only do you need to
- 1:33know the 90
- 1:34level of confidence you also need to
- 1:36know the degrees of freedom and the
- 1:38degrees of freedom is going to be n
- 1:40minus 1 but we don't know what that's
- 1:41going to be without knowing n so how
- 1:43would we determine an n
- 1:45similarly you don't know what your
- 1:46sample standard deviation is going to be
- 1:48until you actually take some samples so
- 1:51instead of that what we could think
- 1:52about we know that another legitimate
- 1:56way to construct a confidence interval
- 1:57and the margin of error is to say all
- 2:00right i can take my sample mean and i
- 2:03can add or subtract
- 2:05a
- 2:06z-score a critical value and this time
- 2:08use the z table where if i multiply that
- 2:10times the true
- 2:12population standard deviation and divide
- 2:15that by the square root of n
- 2:18now you might say well i don't know the
- 2:19true population standard deviation but
- 2:21they tell us a pilot study suggests that
- 2:24the driving ranges for this type of
- 2:26vehicle have a standard deviation of 15
- 2:28kilometers so we could use this as an
- 2:30estimate of our true population standard
- 2:32deviation so this is 15 kilometers right
- 2:36over here and then the good thing about
- 2:38a z table is you don't have to think
- 2:40about degrees of freedom you could just
- 2:42look up your confidence interval and so
- 2:44then we could just say
- 2:46that look z star
- 2:49times 15 kilometers
- 2:51over our square root of n
- 2:53this right over here is our margin of
- 2:55error
- 2:56this right over here is our margin of
- 2:57error that has to be no more than 10
- 2:59kilometers
- 3:01so that has to be a less than or equal
- 3:03to
- 3:0410.
- 3:05and we can figure out what z star needs
- 3:07to be for a 90 percent confidence level
- 3:09and so then we just solve for n so let's
- 3:11do that
- 3:12now to figure out z star i could use a z
- 3:14table but just for diversity of
- 3:16methodology let's use a a calculator
- 3:19here
- 3:20so to figure out the z value that would
- 3:22give us a 90 confidence interval i can
- 3:24use a function called
- 3:26inverse norm and you can see that right
- 3:28over here that's choice 3. let me just
- 3:30select that and what it'll do is
- 3:34you give it the area that you want under
- 3:35a normal curve you can even specify the
- 3:38mean and the standard deviation although
- 3:39you want the mean to be 0 and the
- 3:41standard deviation to be 1 if you really
- 3:43want to figure out a z-score here and so
- 3:46it'll do is it will give you the z-score
- 3:49that will give you that corresponding
- 3:50area
- 3:51and so i want and actually it's already
- 3:53selected that i want the center
- 3:55area to be 90
- 3:57so i could say point
- 4:00right over here if i use the left tail
- 4:03then that means if i have 90 percent of
- 4:05the center that means i would have five
- 4:06percent of either tail so instead of
- 4:07doing it .9 and center i could have done
- 4:10.05 and used the left tail or used .05
- 4:14and use the right tail but this is
- 4:16exactly what i want so let me just go
- 4:18and paste this
- 4:20and so this should give me the
- 4:23appropriate
- 4:24z value
- 4:27so there you go if i want this middle
- 4:29ninety percent the center ninety percent
- 4:31i have to go one point roughly one point
- 4:34six four
- 4:36five standard deviations below the mean
- 4:37and that same amount above the mean so
- 4:39it's roughly our critical value here is
- 4:42approximately 1.6
- 4:44let's just say
- 4:451.645 so we have
- 4:491.645
- 4:52times
- 4:54times 15 over the square root of n is
- 4:57going to be less than or equal to 10.
- 5:00and so now there's a couple of ways that
- 5:02you could do this we could do a little
- 5:03bit of algebra to simplify this
- 5:05inequality i encourage you to do so or
- 5:07you could even try out some values here
- 5:09and see which of these ends would make
- 5:11this true and we want the smallest
- 5:13possible one i'll do it the algebraic
- 5:15way because if you're actually doing
- 5:17this in the real world nadia would not
- 5:19have a multiple choice right over here
- 5:20she'd have to figure out the sample size
- 5:22in order to conduct her study
- 5:24so let's do that
- 5:26so let's see if i divide both sides by
- 5:291.645 and 15 what do i get i get 1 over
- 5:34the square root of n
- 5:36is less than or equal to
- 5:3910
- 5:41over
- 5:421.645
- 5:46over
- 5:47over
- 5:4815.
- 5:50and then if i take the reciprocal of
- 5:51both sides i get the square root of n
- 5:54is greater than or equal to from taking
- 5:56the reciprocal of both sides and so this
- 5:58is going to be
- 6:021.645
- 6:04times 15 times
- 6:0815 all of that over 10
- 6:10all of that over 10. see 15 over 10 is
- 6:13just 1.5 so let me just write that as
- 6:151.5 right over here and then if i square
- 6:18both sides i would get that n needs to
- 6:22be greater than or equal to
- 6:261.645
- 6:28times
- 6:30times 1.5 and then all of that squared i
- 6:34just squared both sides all of that
- 6:36squared and so let's get our calculator
- 6:37back we are going to have 1.645
- 6:43times
- 6:441.5 and then we want to square it and we
- 6:48get 6 point approximately 6.0
- 6:526.09
- 6:56so n
- 6:57has to be greater than or equal to
- 6:596.09 and of course our sample size needs
- 7:02to actually be a whole number so what's
- 7:04the smallest whole number that is larger
- 7:06than 6.09 well that's going to be 7. so
- 7:10that would be this choice right over
- 7:12here this is the smallest approximate
- 7:14sample size required to obtain the
- 7:16desired margin of error and of course we
- 7:19won't really really know until we
- 7:21actually conduct the study we obviously
- 7:23here use an estimate of the population
- 7:26standard deviation we used and we used a
- 7:30z table but it will be interesting when
- 7:33nadia actually conducts the study to see
- 7:34if her margin of error is indeed no more
- 7:37than 10 kilometers with a 90 level of
- 7:40confidence
About this transcript
This page contains the full transcript of Sample size for a given margin of error for a mean | AP Statistics | Khan Academy by Khan Academy, generated from the public captions YouTube serves with the video. The transcript has 1,271 words across 204 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.