Fire The Grid |
|
The "Fire The Grid" event set for July 17 2007 brought a huge number, perhaps 10 times as many as for similar events, of notes from around the world, interested in notifying me of it and asking whether we would look at it as a GCP event. We decided early that it would be included in the formal series, and specified exactly the time defined by Shelley Yates in her presentation and descriptions. Here is an example email, which has the necessary information and some interpretations that help explain how the event is of importance:
The formal event was defined as the one hour period from 11:11 to 12:11 GMT on July 17. The Chisquare is 3731.7 on 3600 df, for p=0.062 and Z=1.542. This is in accord with the prediction, and is marginally significant. Below, we provide exploratory analyses that give some context for the formal analysis, since so many people are interested. It is important, however to understand that single events are not expected to answer the subtle questions we ask in the GCP. Because the average effect size is small (Z ~ 0.3) we cannot generally interpret single events with reliability; we need repeated tests of the same type of hypothesis. That having been said, we see that the Fire the Grid event has a positive outcome, consistent with the prediction, and indeed shows a larger deviation than the average. I encourage people interested in this and other occasions that bring us together to meditate or pray for a better future to trust your/our feelings and direct perceptions of the effects. The GCP analyses are a very interesting adjunct, but they should not be seen as a test of the intent or success of the gathering of minds for good purpose.
The full day of July 17 (UTC) is shown in the next graph, with the hour-long Fire the Grid (FTG) event marked. While the cautions about interpretation of single analyses apply here also, it is interesting to note that the FTG presents a notable "spike" of positive trend on a rather consistent day-long background trend that is negative. For a larger context it is instructive to look at a long term trend analysis that shows some suggestive correlation with polling data that in turn may reflect our communal state of mind. Here too we see a general trend that if regarded as the "true" expectation would make positive deviations more impressive.
Again, we do not believe it is appropriate to interpret single events even though it is tempting to do so when the "picture" looks good. The signal to noise ratio is way too small, and to make claims based on any single experiment in this domain is potentially misleading. It happens in this case that the FTG event is one in a decent size collection we have assembled over the years, and it is consistent with the general outcome. Good stuff, but just the beginning of learning about the stuff. Numbers EngagedWe have modestly persuasive evidence that the effects in GCP data are a function of the numbers of people engaged. When millions are likely aware of an event and likely participatory in some sense, the effects are larger and more reliable. We don't have an objective count and have to rely on estimates and best guesses for the most part. The Fire the Grid event provided an opportunity to look at web-based counts, as a step toward a useful measure of N. Here is what Google and YouTube looked like over the time around the FTG event.
|