I love reading literature but when discussing results with fellow clinicians, I feel that there is a knowledge gap in understanding what is significant and what is not. This is likely due to the wide range of statistical analysis used to determine the results in different types of studies. I plan on developing a series of posts to help PTs understand the basics of stastical analysis, clinical significance as well as the basic research design.
When designing a research project, the choice of the specific design depends on six critical questions.
- How many independent variables are being tested?
- How many levels does each independent variable have?
- How many groups of subjects are being tested?
- How will subjects be selected and how are they assigned to groups?
- How often will observations be made?
- What is the temporal sequence of interventions and measurements?
Review: The independent variable is the variable that is presumed to cause or determine a dependent variable. It is manipulated or controlled by the researcher.
From these questions, a researcher will determine the type of design. This list is not exhaustive but highlights some commonly used strategies for the construction of a true experimental design (exerts control over threats to internal validity). A true experimental study is one in which subjects are assigned to at least two comparison groups.
Single-Factor, True Experimental Designs (One Independent variable)
- Pre-test – Post-test control group: this is a basic study that is used to compare two or more groups that are formed by random assignment. Often, one group recieves the experimental variable and other does not. It is used to establish a cause and effect relationship. It is better known as a randomized controlled trial or RCT.
- Post-test only design: identical but the pre-test is not administered to either group. This is used if the pretest may be impractical or potentially reactive.
Multi-Factor, True Experimental Designs (>One Independent variable)
- Factorial Design: incorporates two or more independent variables, with independent groups of subjects randomly assigned to various combinations of levels of the two variables. These studies are often denoted accoring to their diminsions of number of factors, so a two-way facor design has two indepndent variables and a three-way would have three. They also can be described by the number of levels within each factor. A 3×3 would indicate 2 variables, each with 3 levels within each factor, and so-on.
- Randomized Block design: this is used when an attribute variable, or blocking variable, is crossed with an active indep variable. This is used if a factor is thought to influence differences between groups.
Foundations of Clinical Research: Applications to Practice 2nd Ed. Portney LG & Watkins MP. Prentice Hall Health. Upper Saddle River, NJ. 2000.