The scientific process follows a distinct path from hypothesis (idea) to established knowledge. It is possible to skip steps and derive directly an outcome but the conclusions are less certain without supporting evidence from reliable experimentation.
The scientific method provides the most reliable way to increase our knowledge. Nevertheless, like other human activities, scientific investigation is subject to potential errors, personal opinions and uncertainties. When weighing up the evidence for potential health effects, scientists consider different aspects before drawing their conclusions.
- Researchers may report preliminary findings at conferences to share knowledge and obtain feedback. The may revise the hypothesis or experiment based on this feedback. Interim results presented at conferences have not benefited from peer review and should be interpreted with caution.
- Established Knowledge:
In order to arrive at a scientific consensus, the results of many experiments are compared through a weight of evidence approach. This is not simply a case of counting publications, individual peer reviewed studies are examined and their strengths and weaknesses noted. This allows overall conclusions to be formed about the strength of the evidence for the hypothesis and a consensus of what is established knowledge about a subject. However, science is not static and even established theories may be challenged by new experimental findings that have followed the scientific method.
Conducting a well-designed experiment is the most important step of the process. Much effort is made at this stage to properly design or “control” the experiment so that other external factors (“confounders”) cannot influence the outcome. The collected data is analysed and the hypothesis must be rejected or modified if not supported by the experimental outcome.
The hypothesis is best described as a tentative theory or a working assumption. It is developed following research to find out what is already known about the observation. The hypothesis is then used to make predictions of probable effects that can be tested by experiment.
Results may be further tested by independent replication. Independent replication is essential to minimising the effects of potential personal and experimental bias. The follow-up studies may try to duplicate the original study (direct replication) or try to improve some aspect of the experimental conditions (confirmation study). If the original results are confirmed then greater weight is given to the hypothesis. Much greater weight is given to the original results.
Identification of a specific idea is always the first step. This may be an original idea, the result of other background research or sometimes based on anecdotal reports but it must be properly identified and clearly defined.
Following acceptance by the journal editor, the paper will be prepared for publication and assigned to a future issue of the journal. Increasingly, accepted research papers are made available on-line before appearing in the print journal. After it appears, other scientists may comment on the results by writing to the journal and the researchers can choose whether to respond. Publication in a assessments use this as a threshold for considering scientific evidence. However, not all published research is without flaws and the quality of peer review varies across journals.
- Peer Review:
During peer review the study design and results are critically examined by one or more reviewers with expertise in the subject. The journal editor selects the reviewers and peer review is usually conducted anonymously. Peer review is important for quality control of research claims and protects the reputation of the researcher and the journal. The reviewer may recommend acceptance, amendment or rejection of a research paper to the editor of the scientific journal. The process of submission, review, amendment and resubmission may take several months.