Minimum sample size
Six Sigma – iSixSigma › Forums › Old Forums › General › Minimum sample size
- This topic has 16 replies, 17 voices, and was last updated 14 years, 1 month ago by
TonyBo.
-
AuthorPosts
-
March 28, 2008 at 10:23 am #49718
Louise PParticipant@Louise-PInclude @Louise-P in your post and this person will
be notified via email.I wonder if you can help?
We are analysing data from an inbound call centre that receives around 25,000 calls per day. To analyse a significant proportion of these, what minimum percentage would you recommend?
At the moment we have 600 calls from about 3 weeks, but feel this sample size is too small?
Thanks in advance, Louise0March 28, 2008 at 4:51 pm #170227I found a very smart system that will guide you perfectly through this as well as the analysis and how to read it, need more info ? I will be glad to help with directing you to the source, it is a life saver. [email protected]
0April 7, 2008 at 10:01 am #170687
Bruno TernonParticipant@Bruno-TernonInclude @Bruno-Ternon in your post and this person will
be notified via email.Could you please direct me to the “source”? Thanks!
Kindly,
Bruno0April 8, 2008 at 5:29 pm #170784I wonder if you can help?
We are analysing data from an inbound call centre that receives around 25,000 calls per day. To analyse a significant proportion of these, what minimum percentage would you recommend?
At the moment we have 600 calls from about 3 weeks, but feel this sample size is too small?
Thanks in advance, Louise—————————-
Could you please direct me to the “source”? Thanks!
Kindly,
Bruno————————————————–
Try this link, call them, they are always there to help.
http://www.mentorsoftpro.dbar-innovations.com/ or
http://www.dbar-innovations.com/
Hope this helps,
Respectfully,
Vince0April 8, 2008 at 5:56 pm #170785Hello Louise,
I have a three page document about how to calculate sample sizes for both Variable Data analysis and Discrete Data analysis. It is part of our black belt training manual.
If you provide me with your email address, I will send it to you.
Helper0April 8, 2008 at 6:05 pm #170786Helper:
Please send a copy of your Sample Size document.
Dave
[email protected]
0April 8, 2008 at 7:51 pm #170793It Depends:
Based on some knowledge of call centers you have a large stratification of calls. You did not state your intended goal so I would 1) Stratify the nature ofthe incoming calls, 2)Take your sample from a proportionate representative sample.
i.e. if you have 10% Technical, 80% informational, 10% Financial use these to select the number of each type to obtain information.
Regarding the number of samples: use the maximum number you can analyze in the time you have to performr this task. More is better.
If you are looking for a minimum number use a simple software package avaialbe to calculate power and size.
0April 9, 2008 at 1:58 am #170810To Louise,
Did the software at Dbar-innovations posted by Vince help you ? Did you try the live test drive and use the analysis tools ? I am curious,
Thanks,
Doug0April 9, 2008 at 2:57 pm #170846
GreenalghParticipant@LouiseInclude @Louise in your post and this person will
be notified via email.Hi Helper thanks for the offer that sounds very useful, my email is [email protected]
0April 9, 2008 at 5:15 pm #170851Please send a copy of your Sample Size document. Thanks
[email protected]0April 10, 2008 at 5:07 am #170894
SwaggertyParticipant@GeorgeInclude @George in your post and this person will
be notified via email.Hi Helper,
Can you please share that we me too? I had posted an article some time back on the same ( in response to another post- I had checked with somebody else, and had come up with 3 different formulas..each of which was right in its own way..but gave different results)
, so i would appreciate a standard formula.
My email is [email protected]
Thanks in adv,
George0April 20, 2008 at 8:36 pm #171335If you could, send me a copy of it, too. My email address is [email protected]. Thanks!
0April 21, 2008 at 6:32 am #171341
Amit JhaParticipant@Amit-JhaInclude @Amit-Jha in your post and this person will
be notified via email.Dear Louise,
The sample size that you want to decide will depend on how minutely do you want to onitor the calls received, in other words what is the margin of error that you are OK to live with, it statistical terms what is the confidence level and confinence interval that you are looking for in your test.
In the given scenario, the thumb rule is to go with 95% confidence level and 98% confidence interval. So you receive 25000*7 calls per week. So you can monitor 2,369 calls per week. If working days are 5 then you may monitor 2356 calls per week.
After a given population size, sample size becomes independent. Therefore you may have seen bevy little decrease in sample size even though thepopulation has reduced majorly.
You may easlity find on the net online sample size calculators. Read some more articles and also learn about the power of test.
Let me know if I can be of any further help.
Amit0April 21, 2008 at 7:45 am #171342
MarshallMember@Rob-MarshallInclude @Rob-Marshall in your post and this person will
be notified via email.If you would be so kind to email the sample size document to me that would be most appreciated
[email protected]0May 12, 2008 at 9:09 pm #171924I tried the link for mentorsoft and it did not work, the link to http://www.dbar-innovations.com did work. When I got there I discovered it was a software system with a DMAIC Mentoring style analysis system and a tracking program. No info on consulting, but interesting system.
0May 17, 2008 at 6:34 pm #172057Hi Helper, When you get a chance could you send it to me as well.
[email protected]0May 18, 2008 at 2:52 am #172061Use a sample size calculator..
0 -
AuthorPosts
The forum ‘General’ is closed to new topics and replies.