If experiments don’t produce positive results they are often not published, yet they can help to progress research. Should we publish them?

$