Scientists should stop using the term 'statistically significant' in their research, researchers urge.