Correlational research methods - Essay Example

According to the study conducted the goal of pursuing correlation research is to identify and describe the relationship between variables and the strength between such relationships. Analysis may involve relationship of more than two variables at the same time. For example, in the study conducted by Dickson-Spillmann and Siegrist on dietary behavior, the authors investigated the prevalence of misconceptions about healthy eating through survey of procedural nutrition knowledge. The questionnaire consisted of 13 questions, nine of which were based on qualitative consumer interviews while the rest came from expert guidelines.

The aim of the study is to determine if a correlation exists between the consumer-respondents knowledge on health diet and dietary behavior. Correlations may be utilized to predict the behavior of one variable based on how another variable behaves. However, even though two variables may be highly correlated, one variable does not necessarily cause the other. Though correlational methods may not be as versatile in determining causality, it does offer several benefits: (1) correlational relationships may provide cues to actual causes; (2) help determine vulnerable groups; and (3) improve comprehension of relationships between several variables.

In Dickson-Spillmann and Siegrist the authors chose to pursue a correlational study approach since the variables cannot be manipulated and the study deals with quantitative variables. Correlational research methods, however, do have its limitations. It doesn’t have the capability of demonstrating a causal relationship between two variables, something which an experimental research method is capable of (Sigelman and Rider, 2009). Pearson correlation coefficients are the most commonly utilized methods.

However, its use requires that the two variables being analyzed are measured in interval or ratio scales, or that the variable are continuous (Urdan, 2001). A correlation or link may be categorized as positive, negative, or zero.  A positive correlation describes variables with values that move in the same direction. As one variable goes up in value, so does the other variable and vice versa (Weiten, 2008). On the other hand, a negative correlation describes variables with values that move in opposite directions.

As one variable goes up in value, the value of the other variable goes down and vice versa (Coon and Mitterer, 2010). Meanwhile, a zero correlation denotes that a linear relationship between two variables does not exist, and, therefore the behavior of two variables does not show any predictability (Spatz 2011, Zechmeister, 2001). In the Dickson-Spillmann and Siegrist (2011) study, the following were observed: (1) respondents who answered majority of the questions correctly ate more vegetables compared to respondents who got a lower number of correct answers (r = 0.29); (2) higher nutritional knowledge is attributed to the female gender, younger age, higher education, nutrition-related qualifications, and not being on a diet program (P < 0.001); and (3) majority of the respondents have the misconception of eating less as a healthy dietary behavior.

Though a correlation exist on several variables, it does not indicate causality. Non-significance of correlation coefficients does not necessarily translate to a lack of relationship between variables. An alternative explanation would be that the relationship is not linear, but curvilinear or cyclical. In other words, if the correlation is not significant, the equation which describes the relationship between two or more variables generates a curve, not a line. This makes correlation coefficients unable to stand on their own as important statistics.

Correlation coefficients attempt to process complex relationships between variables into a