Within certain limits, slime mold is hypothesized to grow at a rate proportional to its size. That is if y(t) = size of slime mold at time t, [(dy)/(dt)] = k y (t), where [(dy)/(dt)] is the rate of growth.

If the size of the slime mold is measured at time intervals of 1 day for 300 days, we have 300 pairs of data. How could linear regression be used to check the significance of this hypothesized growth model?