In an experiment, a farm with infected plants is subjected to a widely applicable insecticide. This insecticide is expected to increase the number of healthy plants after its application. However, at the end of the experiment, the proportion of healthy and infected plants remained the same. Here, the null hypothesis that the insecticide has no effect seems to hold, but should one accept the hypothesis or fail to reject it? Accepting this hypothesis would mean that the insecticide is ineffective and cannot improve the plants' health. This decision actually overlooks the other plausible explanations for the observed results. In this case, using an unprescribed amount or concentration of insecticide might have resulted in no effect. There is a possibility of plants being infected by something that the insecticide cannot target. Failing to reject a null hypothesis means there is no sufficient evidence for the expected or the observed effect. Today, if scientists had accepted null hypotheses, the discovery of plant viruses or the rediscovery of many extinct species would not have been possible.