Research Questions are concrete, are measurable, and answer a pressing research need.
Steps in Developing a Research Question
1) Define: define the broad area you want to study (for example, "private prison efficacy," "alternatives to use of police force on the mentally ill," "prison educational opportunities and recidivism")
2) Describe: describe the problem in simple terms (for example, "X number of lethal force incidents in Y community involve mentally ill individuals undergoing a psychotic break.)
3) Specify: specify what we (the larger research community) don't know about this topic (for example, "Does the creation of specially trained police units to deal with mental health problems decrease lethal force involving individuals experiencing a psychotic break?" or "Are strategies adopted by larger municipalities able to be scaled down for smaller populations with the same results?). Too broad a question will hinder you in being able narrow down to a single answerable question, too narrow a question will hinder you in finding enough research to answer.
4) Create: create your research question based on the information you've gathered. Make sure your question is concrete, specific, measurable, and answerable.
You may go between these three steps multiple times before hitting on a good research question. Don't be discouraged if you try a few dead ends before you have gathered enough information to write your research question.