Michael Jackson Dies
The news shows in the US, especially cable, were preoccupied with Jackson's death and were doing histories and retrospectives on his life as one of the most popular pop stars in the world. A number of correspondents proposed that the death of Michael Jackson should be a "global event" to which our network might respond. Several were willing to make independent predictions about the amount of time we should consider. The majority suggested a full day, which is what we have used. Shorter periods can be assessed informally from the cumulative deviation history of the data.
The GCP hypothesis test is for the 24 hour period beginning at 14:00 PDT, which is some 26 minutes before the time of death given by Michael Jackson's brother. The result is Chisquare 86825.88 on 86400 df, for p = 0.153 and Z = 1.024. The graph shows a fairly steady slope with a modest terminal p-value conforming to the prediction.
We also set the Memorial as a formal event.
It is important to keep in mind that we have only a tiny statistical effect, so that it is always hard to distinguish signal from noise. This means that every "success" in our predictions might be largely driven by chance, and every "null" might include a real signal overwhelmed by noise. In the long run, a real effect can be identified only by patiently accumulating replications of similar analyses.