BookmarkSubscribeRSS Feed
Choose Language Hide Translation Bar
Community Trekker

Data Analysis with a Small Data Set

Hello everyone,

I have a small dataset (only 6 samples). Of those 6 samples, one of them is out of specification in regard to a Y-variable. I want to understand how much my X-variables are contributing to the sample being out of specification (if they indeed are) but I'm not sure what kind of analysis to perform on such a small data-set.

Any ideas?

Thank you so much!

0 Kudos
Super User

Re: Data Analysis with a Small Data Set

Graph Builder...if you see anything that "looks interesting" then design an experiment (data collection plan) for gathering additional data to confirm or refute any "interesting findings" you found in your graph builder analysis of your 6 points.


Re: Data Analysis with a Small Data Set

To reinforce my former colleague @KarenC 's sage advice...all good analysis starts with what I call my three rules of successfull data analysis:


1. plot the data.

2. Plot The Data.



Can you share a bit more information about the practical problem at hand and the methodological approach to gathering the six observations? Is the response(s) numeric or categorical? Did you conduct a designed experiment? If so can you share the design? If not a DOE, then can you describe how the data was actually collected? What do you know about the measurement systems used for the responses? Were they in control and operating according to specifications? Maybe your one suspicious response is measurement system induced? Last thing you want to do is start making causal influences with an unknown or perhaps dysfunctional measurement system.