What Is The Next Step If An Experiment Fails To Confirm Your Hypothesis?

The most common framework used when performing an experiment is the Scientific Method. The hallmarks of the Scientific Method include: asking a specific question, devising a hypothesis, experimenting to gather data, analyzing the data, and then evaluating whether the hypothesis is correct based on the experimental data. When the data support the hypothesis, the findings can be published or shared. However, what happens if the findings do not confirm the hypothesis? Here are possible next steps to take.

The write-up is part of the evaluation process of the experiment. No matter what happened during the experiment, the results have to be shared, whether they confirm or deny the hypothesis. Assess all stages of the experiment – the hypothesis, the experimental stage and the analysis phase – and disclose the results. Next, identify problems that arose during the experimental process and follow that in the write-up with suggestions for improvements and future courses of action. The key to crafting the section on future courses of action is to work systematically backward to ascertain where the error might have taken place and then to make corrections to see if changes in those gap areas might lead to different results. The write-up is necessary to document what happened during the experiment. It becomes part of the background literature surrounding the issue being questioned or experimented on.

Make slight changes in the process by methodically working backward, starting with a check on the analysis process. Was the analysis off? Sometimes experimental data are incorrectly assessed. That means you have to ascertain if the analysis is where the error lies. For example, some physics experiments require mathematical calculations. If these calculations contain errors, then the analysis shows data that does not coincide with the hypothesis. Correcting any mathematical calculations is a necessary step after any experiment, especially if they have a bearing on whether the data confirms the hypothesis. Besides mathematical calculation analyses, evaluations that center on comparisons, making predictions or making discoveries can occur. If analyses reveal discrepancies, check whether there were any errors in the comparisons, predictions or discoveries processes. Rooting out these errors can alleviate any data-to-hypothesis discrepancies.

Human error can skew experimental data, and human error can rear its head at the experimental stage – whether in setting up the experiment, running the experiment, observing the experiment or in tabulating the experimental results. Minimizing errors at the experimental stage can affect whether the results confirm the hypothesis or not. There may have been other variables that arose that were not anticipated or could not be measured that affected experimental results.

Perhaps a different experiment can better test the hypothesis. There are situations in which an experiment is not the appropriate type to test a hypothesis. Perhaps design problems arose that were not evident in theory or on paper but became apparent during actual application. If so, an entirely different experiment may be needed. Experiments are essentially approaches and data-gathering methodologies to test a hypothesis. In other words, Experiment A utilizes Approach/Methodology A to test the hypothesis. If the results do not confirm the hypothesis, devise Experiment B with Approach/Methodology B.

If several different experiments all reveal that the hypothesis has not been confirmed, a revision of the hypothesis is in order. Perhaps it was the hypothesis all along that needed amendment. If so, devise a new way to ask a question and formulate an educated guess. Was there something amiss in the cause-and-effect relationship? Were associations and correlations assumed incorrectly? Remember that a hypothesis is a tentative description of some phenomenon. If several reproducible experiments show the hypothesis does not work, then it might be time to reject the hypothesis and replace it with a more viable one.