I have a hypothesis!

Once you read this post, you will have a better understanding of how to write a hypothesis or hypotheses (plural). The previous statement is an example of a prediction I have made about your ability to read and gain knowledge, and hypotheses are predictions. Most of the time, we start with a question and then create a hypothesis based on the question. Occasionally, you can work backward and create a hypothesis first. Either way, you need a hypothesis to test.

One exciting aspect of research is we create hypotheses we believe will be right, but sometimes we are wrong. Science! We often base our predictions on previous studies, anecdotal evidence, or both. When there is no or very little evidence to base a hypothesis on, researchers make educated guesses based on experiences. Think about how little information existed for Florence Nightingale to create a hypothesis around the importance of handwashing.

A more recent example would be the Coronavirus pandemic. The global hypothesis for decreasing the spread of the disease was to social isolate, wash hands, and wear masks. However, this prediction was based on results related to other conditions about which we know much more. Suspend current knowledge momentarily, and imagine writing some hypotheses about Covid-19 in March of 2020. The hypotheses might look something like this:

1.?????? The rates of Covid-19 related deaths will be less in states who followed strict guidelines, including social distancing, wearing face masks, and stay at home orders for greater than three months.

2.?????? Patients with pre-existing heart disease will have increased rates of Covid-19 related deaths.

3.?????? Pediatric patients between the ages of 1-10 years will have a lower incidence of Coronavirus infection.

Maybe you noticed, the hypotheses have directionality, i.e., increase, less, decrease and provide an explanation by way of age, co-morbidities, or actions. What a hypothesis does not look like is:

1.?????? Covid-19 deaths will get worse before they get better.

a.?????? Although it is a prediction, it lacks qualifiers like direction, location or population

2.?????? We will have the Coronavirus under control in a few months.

a.?????? Too vague, there are no measures or way to test the hypotheses

Hypotheses can also be non-directional and speak to the relationship between two variables.

1.?????? There is a relationship between Covid-19 and socioeconomics.

2.?????? There is a relationship between the type of PPE worn and the feeling of safety among healthcare workers.

Finally, there is the null-hypothesis. The null states there is no relationship between variables.

1.?????? There is no relationship between Covid-19 and socioeconomics.

2.?????? There is no relationship between the type of PPE worn and the feeling of safety among healthcare workers.

If you have an idea for a research question - get started today . . . make predictions and test them!

_____________________________________________________________________________________

Heather Watson, Ph.D., R.N. is the Nurse Scientist for the Johns Hopkins Health System. In this role, she leads and conducts interprofessional research to strengthen the implementation, adoption, and sustainability of best evidence for clinical, administrative, and nursing practice.