The transfer characteristics are generated as figure using the procedure. Each operating point is then identified and a tangent is drawn ateach point to best reflect the slope of the transfer curve in the region. An appropriate increament is then choosen for $V_{GS}$ to reflect a variation to either side of each Q-point. Equation is then applied to determine $g_m$.

$$1. \ \ g_m=\dfrac{\triangle I_D}{\triangle V_{GS}}\cong \dfrac{2.1 mA}{0.6V}=3.25mS \\2. \ \ g_m=\dfrac{\triangle I_D}{\triangle V_{GS}}\cong \dfrac{1.8mA}{0.7V} \cong2.57mS \\ 3. \ \ g_m=\dfrac{\triangle I_D}{\triangle V_{GS}}= \dfrac{1.5mA}{1.0V}=1.5mS$$

The decrease in $g_m$ as $V_{GS}$ approaches $V_P$.

In transfer characteristics we have seen that the relation between the drain current $I_D$ and gate to source voltage $V_{GS}$ is non-linear. The relationship is defined by Schockleyâ€™s equation.

$$I_D=I_{DSS}\bigg(1-\dfrac{V_{GS}}{V_P}\bigg)^2$$