First issue will be, what is "a value close enough" to the "true" value? Not a simple matter, as you can surmise.

If you do a one-way analysis of variance, instead of a mess of t tests, you will pull out a generalized variation in counts, which is to say, a standard deviation (stdev).

If you do it with the 3 samples & 20 repetitions as you suggest below, I _think_ you will wind up with an estimated stdev for the sample counts. And this _assumes_ that the samples were pulled independently from the same flask, and that the 20 count locations were not related to one another in any way.

Then there is the question of the distribution of particle counts. It could be well away from a "Normal' dist, which might throw off your interpretation of what a stdev means. Have you checked?

Let's assume for a moment that the dist. is Normal, and proper independence exists. We still need to know what 'close enough' is.

If you could say "an estimate of mean within 'delta' is close enough" then I would suggest that you use


n ~>= (2*s/delta)^2


where s is the stdev.

I'm sure this estimate can be refined, and should be. No matter what you do, however, you have to say what 'close enough' is.

Cheers,

Jay

Euh wrote:

Hello all,

I'm trying to evaluate the concentration of suspended particles �n a
flask.
I took 3 differents samples from the flask and, for each sample, I did
20 counts under the microscope (= at 20 different locations on the
microscope slide)

I ended up with the data posted below this post.

My question is:

Given this data, what is the minimum number of samples and/or counts
required to insure that I get a value close enough to the true value ?

To adress the issue of "how many samples", I was thinking of
compairing the mean of each column with the global mean using a t-test
(since the variance is unknown). Is this the right approach ? Should I
use the paired or unpaired t-test ?

To determine the minimum number of counts, what should I do ?

Thanks for the help

Data (each column is a sample, each row is a count)

98      108     115
123     102     120
88      90      93
65      71      92
95      56      131
68      145     138
114     136     116
82      100     98
87      85      70
109     116     56
134     157     34
102     60      113
130     90      125
53      121     63
124     111     81
114     92      117
118     131     109
76      55      113
97      256     108
79      146     72
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================





-- 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, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================

Reply via email to