Relationships Between Variables
The goal of any research study is to examine the relationship between the independent and the dependent variable(s). The variable that causes or influences the change is called the independent variable, and the variable that is measured for change is the dependent variable or outcome. Independent variables are controlled by the researcher. A good way to learn the difference between the variables is that the dependent variable depends on the other, and the independent variable is independent of change.
The type of study used to examine this relationship determines how much one can learn about the relationship. To examine relationship, one must first examine descriptive data. With descriptive data, the researcher aims to determine whether a certain condition exists and, in terms of epidemiology, its prevalence. For example, a researcher reviews injury data and identifies how many cases of a particular injury occurred in a group of athletes.
After the researcher collects descriptive data, he or she may want to examine the possible relationship between variables. They may be correlated or associated. When 2 variables are correlated, they are occurring at the same time and are linked together. Figure 3-1 shows 2 overlapping circles. Each circle represents a variable. The amount of overlap can be measured; the greater the overlap, the more strongly the 2 variables are associated with each other. Association does not mean causation; it means that 2 things are related and could impact each other. Cross-sectional studies are an example of a study design that is made to examine the association between 2 or more variables. In this type of study, the researcher only measures variables once, instead of multiple measures over a period of time. Imagine taking height, weight, and girth measurements on different athletes and then asking them how many days a week they exercise. The mathematical relationship between weight and number of days a week they exercise can be assessed by examining the data. They may have no relationship at all or be strongly related based on the value of the correlation coefficient (from 0.00 to 1.00). Even if they are strongly related (values of .7 or greater), the researcher cannot say exercise causes lower or higher weight, but he or she can just that they are correlated.
Weight Exercises For Weight Loss Photo Gallery
The second relationship has the most influence on clinical decision making. It is called cause and effect, or causal. This relationship is harder to establish because the researcher has to control for all other possible reasons a specific outcome might occur. A cause-and-effect relationship is one in which one variable causes a change in the other variable (Figure 3-2). When an athletic trainer applies ice to an acute injury, he or she is trying to change something about the athlete’s condition. The ice is the independent variable, and clinical measures, such as pain or range of motion (ROM), are the dependent variables.
In level 4, a longitudinal cohort study is stronger than a cross-sectional study. A study with 2 nonrandomized groups provides better cause-and-effect evidence than a single group pre- to postdesign in the quasiexperimental studies.
The top level is for systematic reviews. Two types of studies are included here. The first is a meta-analysis, which combines the data of several RCTs to determine the effect size a particular independent variable has on a dependent variable. The second, called a systematic review, is a position paper. These are detailed literature reviews conducted by a group of experts that determines what the standard of care should be for clinicians in their field. The National Athletic Trainers’ Association has produced several position papers in the past decade that set the standard of care for athletic trainers. Other medical organizations, such as the American Medical Association and American Physical Therapy Association, also develop practice guidelines with position papers. Often in position papers, the authors use another system besides the evidence pyramid to grade the support for practice standards. One such system is called the Strength of Recommendations Taxonomy. This system considers the strength of the study design and the type of dependent variables measured in the study. Measures that are closely tied to patient outcomes, such as changes in symptoms, level of function, cost, or quality of life, are given more weight than disease markers and risk factors. Using a grid (Table 3-1), the Strength of Recommendation Taxonomy grades the evidence as A, B, or C.