Table 2.

r2 values for best fit of normal probability distributions

LognormalCubed rootUntransformed
Controls0.9870.8310.318
0.5 μm quinpirole0.9980.9510.680
40 μm sulpiride0.9960.9050.462
Quinpirole/sulpiride0.9970.9490.688
50 μm l-DOPA0.9940.8800.416
Quinpirole/l-DOPA0.9970.9250.553
  • Linear best fit regressions were calculated for normal probability distributions of raw values and cubed root and log transformations of the quantal sizes. The population distributions are shown in Figures 5 and 6. In each case, the data points were better fit by a lognormal transformation than a cubed root transformation.