Does the sign of NOV/s Arctic Oscillation index (AO) of -0.530 have any predictive power for the sign of this winter/s AO?
The 2x2 contingency table below suggests it does.
At the 95% confidence level ... the Chi-square 'Test of Independence' shows the negative sign of NOV/s AO and the AO sign for D-J-F are not independent but are related. There/s a 2.8% probability the test indicates falsely they are not related.
Over the AO/s 64 year period-of-record ... the data shows when NOV is negative ... the winter/s AO will be also negative.
There/s also a statistically significant relationship between the negative sign of NOV/s AO and the sign of DEC/s AO.
p = 0.028
|True -||76%||50%||True +|
|False -||50%||24%||False +|