How do I integrate the results of these three experiments to generate a final z-score or p-value? Similarly, suppose my three z-scores were all 'almost' significant (e.g. 1.9) - could integrating the results of the three trials result in a rejection of the null hypothesis?