A A A

I wonder how we can design a study that minimizes threats to internal validity?