Hypothesis statements are developed primarily to provide a basis for statistical analysis. The null hypothesis states that there is no difference. In a study comparing a medication and placebo for treatment of a disease, the null hypothesis would predict no difference in measurable effect between the two interventions. Hypothesis statements are not required in descriptive studies. 

A helpful learning resource is the National Institute of Allergy and Infectious Diseases primer on generation of a hypothesis statement.

Also see:

• Establishing Timelines
• Evaluating the Existing Evidence
• Developing the Research Question, Part I
• Developing the Reseach Question, Part II
• Efficacy vs. Effectiveness
• Writing Specific Aims
• Data Sources
• Study Design
• Survey Research
• Gaining Institutional Support
• Biostatistics
• Human Subjects Protections, Part I
• Human Subjects Protections, Part II
• Data Collection, Display and Manipulation
• Grant Submissions, Part I
• Grant Submissions, Part II
• Grant Submissions, Part III 
• Research Presentations
• Submitting a Manuscript