The foundation of any scientific investigation is its experimental design - a logical outline that guides the gathering and evaluation of information. It is the researcher's plan for testing the validity of a hypothesis.

Much thought and hard work accompany the development of a hypothesis and experiments that will yield clear results. A scientist's ability to ask key questions and to formulate them into testable hypotheses may largely determine the success or failure of a given research project.
The very nature of some questions requires estimations or assumptions to be made. Whenever estimations or assumptions are part of the experimental design, they must be clearly stated and justified as part of the experiment.

• An estimation is based on some type of data. The data must be shown as well as the calculations that lead to the final estimate.

• An assumption is based on some fact.
   
  • An assumption is made like this − "If this fact is true, then THIS fact must also be true."
     
  • Any assumptions that are key to the experimental design must be stated as well as the reason for starting with this assumption.

1. Proper controls must be incorporated into each experiment. A control group receives the same treatment as the experimental group except that the factor being tested is applied to the experimental group only, not to the control.
    
2. Experiments must be repeated enough times to allow comparisons between experimental and control groups. It is through such repetition that data can be compared statistically and a high degree of accuracy obtained.
    
3. Experiments must be designed to avoid bias. A researcher must strive to prevent personal opinion about a hypothesis from influencing how tests are made and must also be aware of the bias that any technique or instrument may introduce in the outcome of an experiment.

• Researchers are meticulous notetakers. They make detailed notes in a record book that becomes a scientific diary of the research project in process. Data, or results, accumulate as tests outlined in the experimental design are completed.
   
• There are two types of data:
   
  • Anecdotal data - relates in words what happens in an experiment − recording experimental observations as well as describing mistakes and unexpected events.
     
  • Numerical data - consists of measurements determined by a person or instrument during an experiment.
   
• Collecting data is a time-consuming, tedious process, and patience is an essential ingredient in science. As results accumulate, the researcher tries to find patterns or relationships in the data.
   
• Intrepreting data by asking critical questions is essential to determining the cause and effect of experimental observations. When the results of repeated tests are consistent and patterns become discernible, the next stage of the process is reporting the results.

• Science is a powerful group activity, providing many opportunities for correcting errors. Researchers formulate their ideas from data analysis, then describe these ideas at seminars and meetings pertaining to their particular fields. It is through presentations like these that the researcher has an opportunity to interact with others in the field and to see how well their work stands up to peer scrutiny. The researcher then decides whether more experiments are needed or whether it is time to publish the results.
   
• Publishing a scientific paper is the next step in the reporting process. Hundreds of professional societies throughout the world publish journals containing articles that describe original research. Work so published is then permanently available to the scientific communitity. Monitoring current developments and searching literature for pertinent information are ongoing aspects of any scientist's work. Most scientists subscribe to several journals and use abstracting services to find pertinent reports in their filed.

  Understanding Scientific Papers