A correlation or an association in statistical analyses denotes a measurement of the association between binary variables. A correlation is indicated by a correlation coefficient denoted by (r). A statistical correlation coefficient spans from a positive one to a negative one(+1<r< –1). A correlation coefficient may also include a zero. A zero (0) correlation shows that there exists no association between the binary variables being compared (Jackson, 2012). An association between two statistical variables denoted by (r<–1) shows a perfect negative association. When (r) is less than the negative one (–1), it implies that a given variable rises, while the analogous variable drops. Contrary, when (r) is less than a positive one (+1), it signifies that the dualistic variables being compared is associated positively. This means that, as the quantity of one variable rises the quantity of the other variable it is compared with also rises. For instance, several studies have confirmed the existence of a positive correlation between the age and height of a child (Johnson & Kuby, 2012).
Confounding Variables
In statistical analyses, a confounding variable is an extra variable which existence influences the statistical variables that are investigated in a research project. A confounding variable is neither independent nor dependent (Walker, 1999). Therefore, the results obtained do not reveal the real association between the statistical variables being studied. Correlational studies try to find the associations that exist between any given variables. For instance, they try to find how X affects Y, and whether X causes Y. Therefore, in the absence of a confounding variable, any change in X is likely to result in a modification in the outcome, which is Y.
The study had several limitations that acted as confounding factors. One confounding variable in this study was acculturation, and the other one was immigration status. These two confounding variables were because of insufficient sample sizes and data in several of the 24 types of research reviewed (Kornides, Kitsantas, Yang, & Villarruel, 2011).
Brief Description of the Study
The first purpose of the review done by Kornides, Kitsantas, Yang, and Villarruel (2011), was to analyze recent literature on the causes related to obesity and overweight among Latin American children. Secondly, the review of the 24 articles was intended to recommend nursing inferences and define the way forward for the forthcoming research. Kornides, Kitsantas, Yang, and Villarruel (2011), reviewed 24 research studies. Their review revealed that several factors were associated with obesity among Latin American children. They found out that diet, a child’s background, genetics, and acculturation were the major factors associated with obesity in children. Kornides, Kitsantas, Yang, and Villarruel (2011), concluded that additional research was required to verify how the factors related to obesity and overweight could be used in the intervention of efforts among the Latin American families (Kornides, Kitsantas, Yang, & Villarruel, 2011).
Correlations in the Study
Several correlations were highlighted in the study done by Kornides, Kitsantas, Yang, and Villarruel (2011). The first correlation was that parental BMI was a strong predictor of a child’s weight. The second correlation was that a child’s activity, diet and background were the indicators of obesity in Latino children (Kornides, Kitsantas, Yang, & Villarruel, 2011). However, the correlations between these variables were conflicting. Thirdly, a correlation between acculturation and body mass index in Latino children was unconvincing (Kornides, Kitsantas, Yang, & Villarruel, 2011).
SLP topic and Hypothesis
The SLP (Speech-Language Pathology) topic is homelessness and literacy. The hypothesis for this study topic would be; homeless families are at higher risk of reduced health and linguistic literacy, as compared to families that live in homes. One correlation in this study would look for the association between homelessness and reduced literacy. Possible confounding factors would be ethnicity and the level of education.
References
- Jackson, S. (2012). Research Methods and Statistics: A Critical Thinking Approach (4Ed.). Belmont, CA, Wadsworth Cengage Learning.
- Johnson, R., & Kuby, P. (2012). Elementary Statistics (11 Ed.). Boston, MA: Brooks/Cole Cengage Learning.
- Kornides, M. L., Kitsantas, P., Yang, Y. T., & Villarruel, A. M. (2011). Factors Associated With Obesity in Latino Children: A Review of the Literature. Hispanic Health Care International, 9(3), 127-136.
- Walker, J. (1999). Statistics in criminal justice: analysis and interpretation. Gaithersburg, Md.: Aspen Publishers.