I-Bias parameter shift--what causes?

[Marvz]

Hi all,

I'm a Transfer  Engineer of a semiconductor company. My job for now is to correlate the devices parameter from one location transferred to another location. I'm having problems with regards to mean shift of I-bias parameter of one of my device. Mean shift is about 16-22 % from my previous reading in originating location whereas the allowable mean shift is only 5 %. what could be the causes of these large shifts on I-bias parameter. my device is a Precision JFET, High Speed, Dual Operational Amplifier. Thanks.