This is a better example than the apples (I hope). This time is their is a n=x provided.
"Jay Warner" <[EMAIL PROTECTED]> wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > "@Home" wrote: > > > I have estimation of mean / confidence level problem with very litte data > > to go on ie no Std Dev, no population infor etc.. > > > > The unknown is the sample "n" - how large should it be? > > The problem is what percentage of apples at a farm have bruises on them? > > Only facts are confidence level 95% and error of estimate +/- 4% - Given > > this how do you calculate n? > > > > There is a formula which dispenses w/the need for a standard deviation (they > > cancel out), when E is expressed as a fraction of the standard deviation. > > ex. > > N = (confidence coefficient * standard deviation / E)^2 > > N = 1.96 * SD / SD/5 > > = 1.96 * 5 > > > > My book gives examples, but no practical knowledge of how to convert an > > error of estimate into a fraction of a standard deviation? > > > > Thus how would +/- 4% be represented as a fraction of one or more standard > > deviations? > > The number of bruised apples in a barrel (or on a farm) is given to you as a > percentage - 10%, 35%, whatever. > > the number of bruised apples in the barrel is given by (% bruised) * (N = > number of apples in the barrel). > > But I don't think that will answer your problem. > > If you say that an apple is either (a) bruised, or (b) not bruised, then you > are working with a binomial distribution. this can be _approximated_ as a > Normal distribution if the total number of bruised apples is greater than 5 (% > * N > 5). > > In case you have not had experience on an apple farm, I can tell you, that in > my experience, N is a very large number, and the total number of bruised apples > will be much more than 5. So, you can use a Normal distribution assumption. > > Now, go back to the equation you have used so far, only with the letters still > in it, not the numbers. What terms do you have the values for, and which ones > do you not have? You certainly don't know n yet. If you know all but n, then > you can solve for n, and calculate it out. > > I think you are still missing something. but I haven't worked it out today, so > what do I know? > > Jay > -- > Jay Warner > Principal Scientist > Warner Consulting, Inc. > 4444 North Green Bay Road > Racine, WI 53404-1216 > USA > > Ph: (262) 634-9100 > FAX: (262) 681-1133 > email: [EMAIL PROTECTED] > web: http://www.a2q.com > > The A2Q Method (tm) -- What do you want to improve today? > > > > > > > ================================================================= > Instructions for joining and leaving this list and remarks about > the problem of INAPPROPRIATE MESSAGES are available at > http://jse.stat.ncsu.edu/ > ================================================================= ================================================================= Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =================================================================
