When a number of ghost hunting teams from around the world all come together at one time for a shared common purpose, could their effort at "paraunity" in this regard possibly give rise to a hypothesized "collective consciousness" effect? This was the basic question guiding an exploratory field random number generator (RNG) experiment that first began back in 2016, when nearly 90 teams observed the recently-designated National Ghost Hunting Day through their active participation in the "World's Largest Ghost Hunt" (WLGH), in which they simultaneously conducted ghost hunts at allegedly haunted sites located in various regions around the world (for background details, see this previously entry in the PRF blog, as well as the article linked here:)
Following the hindering setback posed last year by the COVID-19 pandemic, the WLGH is set to resume again on National Ghost Hunting Day (which falls on September 25th of the current calendar year), and as part of the on-going experimental effort to explore this event, the parameters and test prediction for the field RNG analysis of this year's event will be specified here in advance. In previous years, the WLGH was held simultaneously during a particular two-hour period of time (10:00 P.M. - 12:00 A.M. Eastern Time, or 02:00 - 04:00 of the next day in UTC), and the general test of the collective consciousness hypothesis was focused around this period. This year, a wider range of time has been allotted to better accommodate the teams that will be participating in the event, with teams being given designated hours throughout the course of the September 25th day, as well as the early morning hours of September 26th, to livestream their activities over the Internet (a full schedule of the livestream hours can be found at this link). In light of this recent change, a modification of the field RNG analytical approach used in this experiment will be necessary in order to make the results of this year's WLGH event comparable with those of previous years. The planned modification can be conceptualized with a helpful analogy to the celebration of New Year's Eve: At each hour of the day on New Year's Eve, the strike of midnight occurs in one of the various time zones spread around the world, and with each passing hour, the strike of midnight shifts westward from one time zone to another, gradually sweeping its way around the globe like a moving wave. As one means of exploring a possible "collective consciousness" effect occurring in conjunction with this kind of moving event, field RNG data can be collected locally from each of these time zones at the very moment that the strike of midnight is occurring there, and then the data from these individual moments can all be statistically combined together into one, so as to simulate a situation as if the strike of midnight was occurring simultaneously all around the world. A conceptually similar approach can be done with the WLGH: Data being collected locally by RNGs running within local geographic proximity of certain haunted sites during a particular designated local livestream hour can be statistically combined with those from other RNGs running in local proximity to other haunt sites at other livestream hours in order to simulate the occurrence of a massively-synchronized group event. The analytical approach can be summarized as follows:
For each designated WLGH livestream hour, any RNGs which are actively running in the worldwide RNG network being maintained by the Global Consciousness Project (GCP) in local geographic proximity to the haunted sites assigned to livestream at that particular hour will be identified. For present purposes, "local geographic proximity" will be defined here as being within the same city, state/province, and/or country. (For sites located within the United States, this will also include the same local time zone, whether Pacific, Mountain, Central, or Eastern.)
Raw data for each designated hour-long WLGH livestream period (in UTC) will be downloaded from the GCP server, and for each hour, the data from each of the identified local RNGs will be extracted.
To fit the data within the previously-defined two-hour formal test period, the data associated with the scheduled WLGH livestream hours will be paired together into two-hour time spans in the following manner (times listed are in Eastern Time): September 25th: 12:00 - 1:59 A.M. 8:00 - 9:59 A.M. 10:00 - 11:59 A.M. 1:00 - 2:59 P.M. 3:00 - 4:59 P.M. 5:00 - 6:59 P.M. 7:00 - 8:59 P.M. 9:00 - 10:59 P.M. 11:00 P.M. - 12:59 A.M. September 26th: 1:00 - 2:59 A.M. These two-hour periods will then be overlapped in the order listed above. This will roughly simulate a simultaneous, two hour-long group WLGH event as was held in previous years.
For each second of the overlapped two-hour period, the data from the local RNGs will be normalized as z-scores using the theoretical mean and standard deviation for a binomial distribution. They will then be combined by taking a Stouffer's Z-score across all the RNGs for that particular second.
The RNG analysis will then proceed as in previous years. Control RNG data, taken one week prior to the WLGH event itself, will be analyzed in the same manner. Following the one initially made in 2017, the main test prediction will be for a significantly negative (or downward) deviation from nominal randomness across all of the combined data from these local RNGs. As a supplemental exploratory hypothesis, it is predicted that the combined data from all the WLGH events that have been held so far will exhibit a negative (or downward) deviation, as well.
My thanks to Maria Schmidt, Michelle Reuss, and Kathrine Sorilos of the WLGH planning staff for their input and assistance in the formulation of this year's prediction.