That idea, asking the right questions, needs to become a research topic for social sciences/business/government/anything that uses statistics (meaning: everything).
As an example, I am working with an organization that is attached to a national level organization. The national level has a survey for the individual organizations to fill out which asks about various aspects of the organization. Filling out this little survey though gives the specific organization I am a part of a pretty bad grade though because this survey fails to ask whether the organization has it’s own building, where the organization is geographically in reference to military and/or minority groups. Pretty important things that will not be taken into account when filing the paperwork I am sure.
Take a look at grant reports for development work world-wide too; they bring to mind a phrase I heard once, “Development projects never fail.” None of the grants will ever report a failure although, if you return to those project sites maybe a year later, it will be awfully difficult to find a success. Take a closer look at the reports though and the things they ask for are “how many individuals are taught x?”; “how many x were built?”; “how much money was made through x-technique?” All factors in a successful development project but certainly not the final factors and, I suspect, not even the most important factors.
This was the strength of journalism: helping show the details lost in the statistics; but, we all know where that discussion goes at the moment.
So, if these questions are unfair, asking the wrong questions or not actually help improve what is going on, it would seem that we need to learn how ask questions properly. What these questions are, I am not sure yet. I will state that the questions are not nearly as easy to ask or measure though looks like we might have to step back from our love affair with numbers or, at least, recognize that they are representations of a much more complicated world. Maybe the first question to ask, what is the specific situation of x in comparison to other places we are measuring?