First, you have to develop a research question. A research question provides the initial information about your research and gives you the necessary base to start off with your work. Then, you have to review the literature which will include the work done by scholars on a particular topic in the past. After that, you will have to develop your own theoretical framework in which you will mention the dependent and independent variables. Besides, you will have to express the relationship between your variables. Then, you will be able to state your statistical hypothesis which can later be tested by analysing the data.
Your statistical hypothesis must contain a statement which you can test with your research. The researchers develop two hypotheses (alternate and null) in order to make the possibilities mutually exclusive and exhaustive. The sign used for alternate hypothesis is H1 and null hypothesis is represented by H0.
You must also know the type of research which you are conducting. For instance, if you are conducting correlation study then your hypothesis will express the correlation between the variables. In contrast, if you are conducting a causal research, your hypothesis must express the causal relationship between the dependent and independent variables.
For example, if you are conducting research on the topic of Motivational Factors and Teacher's Performance in Higher Education and your independent variables include financial incentives and opportunities for higher education while your dependent variable is teacher’s performance in the classroom. Your alternate hypothesis can be ‘Financial incentives leads to higher performance of teachers in the classrooms’. Contrarily, your null hypothesis can be ‘Financial incentives leads to decrease in the performance of teachers in the classroom’.