No, he has a methodology that is facially sound but fundamentally unserious: his team explores a gigantic space of possible experiments that would demonstrate an effect and reports only on those that support his claim. As a result, his results don't reliably replicate. He's using heavily biased statistics, just on a higher order.