Dear Nosheen Qasim

There could be many reasons, all highly depending on your method of assessment. If you use the standard relation between saturation drain current versus the difference between the gate source voltage and the theorertical value squared only the math may already produce a serious error when dividing by a quantity near to zero.

But there appears a more serious issue and that is that for negative and positive values the slope is different. This suggests a changeover of mechanisms. But to assess that one would need more information on the actual method employed.

Ger