Research, Step 2: The Prediction

Research, Step 2: The Prediction


This five-part series details individual steps in effective research—useful in answering complicated questions, pinpointing the best solution to a problem, or creating a detailed analysis of an issue. When assigning a research project, explain all the steps to your students.

Once the central question or “problem” of a research project is defined, the next step is to come up with a well-reasoned prediction, or hypothesis, about what the research will reveal. There are two essential elements to making this prediction:

  1. Base it on what you already know about the topic.
  2. Be prepared to prove yourself wrong.

As a teacher, don’t be too quick to “correct” students whose initial predictions sound outrageous to you. Let them learn by experience where their opinions need adjusting—if they do. School kids, with fewer preconceived notions and greater willingness to upset established order, actually have an advantage over more experienced researchers who are prone to unconsciously favoring results that match their assumptions. Most paradigm-shifting discoveries are made by those relatively new in their fields.

Newcomers aren’t immune to preconceived-notion bias, however, so warn students about the perils of making up one’s mind and then being unwilling to see any contradictory evidence. Explain that learning as you go is part of effective research and that whatever the final results, everyone who manages the process thoroughly and with an open mind is considered successful. Whenever you can, assign projects to groups to ensure a variety of opinions and input.


The other rule for forming a hypothesis: Take an If–Then–Because approach.

Every hypothesis comprises assumptions, reasons, and consequences, whether the format is experimental (“If I drop this ball on the ground, then it will bounce, because the rubber it’s made of has elastic properties”) or statistical (“If I review family histories in this city over the past fifty years, then I’ll find that children who grew up in two-parent families have higher-paying jobs as adults, because stable homes with examples of regular adult cooperation are most conducive to nurturing dependability and resourcefulness”).

“If–then–because” doesn’t have to be written out in those exact words, but the hypothesis must include all those elements. A common mistake is to leave out the “because” on the assumption “everyone already knows that’s true.” You can take the law of gravity for granted in many cases (it’s pretty thoroughly proven through science and experience)—but you can’t hang your actual hypothesis on it unless you’re prepared either to explain how gravity works or to consider the possibility of exceptions. That holds double for assumptions with philosophical or moral implications, where people are especially terrified for their pride (if not their souls) should anything contradict the principles they’ve based their lives on. So when explaining the concept of hypotheses, emphasize the importance of minimizing personal emotional stakes.

Once predictions are made, the fun part begins: the actual research procedure.