Levels of Evidence

The following are the seven levels of evidence, according to the Oxford Centre for Evidence-Based Medicine:

1. Randomized controlled trials (RCTs): RCTs are the highest level of evidence, as they provide the strongest evidence for a causal relationship between a treatment and an outcome.
2. Systematic reviews of RCTs: Systematic reviews of RCTs are also a high level of evidence, as they synthesize the findings of multiple RCTs to provide a more comprehensive overview of the evidence.
3. Other individual controlled trials: Other individual controlled trials, such as cohort studies and case-control studies, are lower levels of evidence than RCTs, but they can still provide valuable information about the effectiveness of a treatment.
4. Case series: Case series are the lowest level of evidence, as they do not provide any control group. However, they can still be useful for generating hypotheses and for identifying potential adverse effects of a treatment.

Examples of Practice Change

The following are some examples of practice change that could result from each level of evidence:

• Level 1 (RCTs): An RCT might find that a new drug is more effective than the standard treatment for a particular disease. This finding could lead to the new drug becoming the standard of care for that disease.
• Level 2 (Systematic reviews of RCTs): A systematic review of RCTs might find that a particular type of surgery is more effective than another type of surgery for a particular condition. This finding could lead to surgeons changing their practice patterns and performing the more effective surgery more often.
• Level 3 (Other individual controlled trials): A cohort study might find that people who eat a certain type of diet are less likely to develop a particular type of cancer. This finding could lead to public health campaigns encouraging people to eat that type of diet.
• Level 4 (Case series): A case series might find that a new drug is associated with a rare side effect. This finding could lead to the drug being further studied to determine whether the side effect is truly caused by the drug and to develop ways to mitigate the risk of the side effect.

Independent, Dependent, and Extraneous Variables

Independent variables are the variables that researchers manipulate in an experiment. Dependent variables are the variables that researchers measure to see how they are affected by the independent variables. Extraneous variables are variables that could influence the dependent variable, but that are not being manipulated by the researchers.

Two Ways to Control Extraneous Variables

There are two main ways to control extraneous variables:

1. Randomization: Randomization is the process of assigning participants to a study group or control group randomly. This helps to ensure that the two groups are similar in all respects except for the independent variable.
2. Matching: Matching is the process of selecting participants for a study group and control group who are similar in terms of extraneous variables. This helps to reduce the potential impact of extraneous variables on the results of the study.

Peer-Reviewed Articles

The following are some peer-reviewed articles that discuss controlling extraneous variables:

• How to Control for Extraneous Variables in Experimental Research by David Howell
• Controlling for Extraneous Variables in Experimental Research: A Review of the Literature by James G. Jaccard
• Statistical Methods for Controlling Extraneous Variables by G. David Garson

Example of Controlling Extraneous Variables

Suppose that a researcher is interested in studying the effects of a new drug on blood pressure. The researcher might assign participants to the drug group or the control group randomly. This would help to ensure that the two groups are similar in all respects except for the drug.

The researcher might also match participants in the drug group and control group based on age, sex, and other factors that could affect blood pressure. This would help to further reduce the potential impact of extraneous variables on the results of the study.

By controlling for extraneous variables, the researcher can be more confident that the results of the study are due to the effects of the drug and not to other factors.