Rameshwar25
Quite Involved in Discussions
Whenever i have done linearity study, my calculated values never satisfied following criteria:
R-Sq to be more than 75%
Zero line to be within confi intervals
ta and tb to be less than t-value (2.001717 for gm=60, alpha=0.05)
can some1 look into my study data attached and tell me what is wrong here?
Can some1 share study data meeting these requirements?
Further, I would say that the examples given in MSA manual for Bias, Linearity, GRR - all are very unrealistic. for example, if we look into bias data given in manual, against reference value of 6.0, the reading vary from 5.6 to 6.4. For the same ref value of 6, how the actual measurements can vary to this extent.
In case of linearity, the readings for ref value 4 vary from 3.8 to 5.1. Is is practically possible to vary to this extent? Even if someone use an instrument with high least count (like measuring tape), this much variation seems impracticable?
I want covers' views on this matter..
Rameshwar
R-Sq to be more than 75%
Zero line to be within confi intervals
ta and tb to be less than t-value (2.001717 for gm=60, alpha=0.05)
can some1 look into my study data attached and tell me what is wrong here?
Can some1 share study data meeting these requirements?
Further, I would say that the examples given in MSA manual for Bias, Linearity, GRR - all are very unrealistic. for example, if we look into bias data given in manual, against reference value of 6.0, the reading vary from 5.6 to 6.4. For the same ref value of 6, how the actual measurements can vary to this extent.
In case of linearity, the readings for ref value 4 vary from 3.8 to 5.1. Is is practically possible to vary to this extent? Even if someone use an instrument with high least count (like measuring tape), this much variation seems impracticable?
I want covers' views on this matter..
Rameshwar
Attachments
Last edited: