How to Write a Hypothesis
Writing a hypothesis for a research question can be challenging and can dictate both the ways in which we conduct our research and how we analyze our results. But before we get there, let's start with the basics: what is a hypothesis?
Although the ways we approach writing hypotheses vary between disciplines; broadly speaking a hypothesis may be seen as: a statement which describes the researcher’s predicted findings. This may be causal (when one thing causes another), for instance “the increase in temperature causes an increase in respiration of plant tissue.” This hypothesis shows a cause-and-effect and is also directional (increasing temperature shows an increase in growth). We also have non-directional hypotheses, where we do not state the direction of movement, for example; “There is a difference between the diets of people living in the UK and France.” This figure explains a positive causal relationship between temperature and plant respiration rate.
There are also non-causal hypotheses, which are more exploratory in nature, for instance “exploring the factors influencing people’s decisions to purchase Nike products.” Here, there could be a multitude of factors and reasons why people make purchases, rather than a clear cause and effect.
So, why write a hypothesis? We want to test our ideas as researchers, but we also want to compare our results with other studies, to back up our own observations, or to refute other bodies of work. Therefore, our hypotheses need to be comparable to others and most importantly, our research needs to be repeatable. In science, we are always contributing to growing bodies of scientific evidence through experimental research.
Dependent and independent variables
Now that we have covered what a hypothesis is, let’s explore a little about how to construct one. Firstly, we need to understand our variables. Variables can be independent (e.g temperature) or dependent (e.g growth rate). If we are conducting an experiment, we will look at how the independent variable affects the dependent variable. This would be a simple hypothesis. More complex hypotheses may include more than one independent and dependent variable. Examples of variables can be found here.
When we write our hypothesis, we want to avoid wordiness. It should be simple, concise and focused. It should also avoid any prior assumptions of the reader’s knowledge in the subject area. For example: “Increased hours of studying contributes to higher exam results.” This is a simple, concise and focused hypothesis.
Null hypothesis
If we are testing our hypothesis statistically, we may want to write a null hypothesis. This is essentially our predicted relationship between two variables where we determine that there is no effect on the dependent variable by the independent variable. For example; H0: There is no effect of fishing intensity on the abundance of fish. Our alternative hypothesis or HA is: Fishing intensity has a negative effect on the abundance of fish. (This is what we actually want to prove, H0 is what we want to refute or reject).
Two types of hypotheses
Largely speaking, we will create our hypothesis before we conduct our research. This is the general method and is called “A priori”. However, if our observations when conducting research make us change our minds or new ideas and concepts come to light, we might want to have more of an exploratory approach to our research and create our hypothesis afterwards. This type of hypothesis is called “A posteriori”.
https://www.slideshare.net/drpsdeb/theory-of-knowledge
In the social sciences, we may have much fewer data points. Perhaps we are exploring human behavior and socio-cultural contexts; in this case, we would predominantly conduct our research using qualitative methods, giving us the ability to conduct in depth research about a small group of people. Just like with a quantitative study, we still need to ensure this kind of research is written so that it is repeatable. For a more detailed look into research questions and hypotheses, read here.
In conclusion, before we embark upon writing our hypothesis, we must first have an idea about what we want to research and how we are planning on doing this. We may even want to have a “why” for the relevance and importance of conducting a particular study. We want to be sure that our hypothesis is testable and repeatable, concise, simple and focused. Once we have all of these in place, we can draft out our statements and then the fun part begins!
Want more support? Reach out to Crimson’s highly skilled team of research mentors at Crimson Research Institute!