We'd established very preliminary evidence that group intention could have an effect on a miniscule biological process. By June 2007, it was time to move up a gear - to send intention to something more representative of a real-life setting. Consciousness researcher Dr. Gary Schwartz and his team at the Laboratory for Advances in Consciousness and Health at the University of Arizona in Tucson. Mark Boccuzzi, our chief lab technician and I began kicking around the idea of sending intention to the seeds of a plant to see if our intention would help it to be healthier than usual.

Mark recommended that we use barley seeds, and we decided to use a simple photo of the target.

I was scheduled to appear before many diverse international audiences during the summer of 2007, which gave me numerous opportunities to test this new experiment in a variety of settings. We scheduled our trial run for late June with an audience of 500 in Australia. Mark had sent me four photos of seeds, each nestled in a seed pocket; during our experiment, I asked a member of the audience to randomly select our target. Our intention: “to enjoy enhanced germination, greater growth and greater health.' After our intention, Mark planted the seeds and during two weeks took regular measurements of each batch.

For weeks I'd advertised our next web Intention Experiment for July 7. But because of our server overload problems I ended up using the NING social network to run this particular experiment. As my regular audience had no advance warning of the new web address, only some 500 people participated.

After a few weeks, Mark wrote me to say his preliminary analysis of the Australian study and our internet study seemed to show no difference between the target seeds and the controls.

Science is a relentless process of correction and revision. The greatest challenge of the scientific method is determining why something works or why it fails. The failure of these experiments could mean that intention cannot make healthy seeds germinate or grow any quicker than normal. Or it could be that intention can't be used to make a 'healthy' system more healthy. Perhaps asking that the plants grow 'faster and be healthier than normal' was not specific enough as an intent.

Or maybe the physical design of our experiment could have 'contaminated' the results: since they were physically planted together, the seeds and germinated plants could have 'shared' the intention via light emissions or even shared dirt. As far fewer people participated than normal, perhaps we didn't have a large enough critical mass of people involved in the experiment to register an effect. Or it could have been another factor we had yet to consider. When venturing out into this kind of strange new territory, the most outlandish possibility must be entertained.

We decided to try the experiment again at a workshop I was holding at Omega in Rhinebeck, New York - this time with a more specific intention: for all the seeds in the target group to sprout at least three inches by the fourth day of growing.' And all the seed samples were isolated from each other, so they would not be able to share 'information'.

A few weeks later, when Gary going over all the Germination Experiment data, he noticed some strange spikes in charts created of the numbers. On closer analysis, he discovered that some 10 per cent of the seeds in each group didn't sprout. From this he realized that the statistical method he'd been using wasn't appropriate to analyze these figures. If the distribution of the figures is not normal, but deviates from a bell shaped curve, the more accurate means of analyzing them is through non-parametric statistics, which don't require normal distribution.

On August 2, he wrote me excitedly to say that he'd come up with some amazing results. For each of the three experiments, the germinated intention seedlings were longer than the control seedlings. Analyzed together, all the intention seeds consistently grew larger than the controls. Gary had used two types of non-parametric statistical methods to analyze a combination of Experiments 1 and 2 because they'd been run identically. The third study had to be analyzed on its own through two statistical methods, as the design was different. According to his calculations, the Omega study had generated significantly larger seed growth overall (in both intention and control seeds) than the other two.

This presented us with even more intriguing possibilities. A group of 500 people scattered around the globe produced the same effect as a group of the same size halfway around the world. According to this data, there was no threshold; even a tiny group of 100 in a room in upstate New York had been able to profoundly affect a batch of seeds 3000 miles away.

Lynne McTaggart is a journalist and the award-winning author of the bestselling book The Field. Her latest book is The Intention Experiment. She also publishes several alternative health and spirituality newsletters. For more information: livingthefield.com & theintentionexperiment.com