Correlations. Correlation analysis Results. Psychology Essay

ESSAY TOPIC: CORRELATION
Your Name
Course no.
Professor
UNIVERSITY
SECTION 1: CORRELATION QUESTIONS
Statistical analysis is performed on numerical data with the objective of investigating the existing relationship between dependent and independent variables (Zou, Tuncali& Silverman, 2013). Thus it is an essential analytical tool used by researchers to gain insight into the nature and type of relationship that exists between variables. Subsequent to statistical analysis, findings and results interpretation is conducted on the dataset. Nonetheless, there exist three types of relationships in bivariate data sets namely positive correlation, negative correlation or nil correlation (Zou, Tuncali & Silverman, 2013). The nature and strength of a statistical correlation is determined using the correlation co-efficient, denoted as r, which depicts the degree of linearity between the different test variables.
The first type of correlation is positive correlation which is also referred to as direct correlation. This type of correlation results when there is a direct proportionality between the test variables that can be depicted using a sloped line graph. This direct mathematical relationship has been described by Huck (2008) as representing “High-high, low-low scenario” (p.49). On the other hand, negative correlation is also referred to as indirect or inverse correlation since it illustrates a scenario where there is inverse proportionality relationship between variables thereby producing a negatively sloped line. According to Huck (2008), the resulting variable relationship can be statistically defined as either “high-low, low-high case” (p.49). Whereas correlation analysis exposes a mathematical relationship, it is unsuitable for making the cause-effect type of arguments. According to Zou, Tuncali& Silverman (2013), research literature misrepresents high correlation to be suggestive of the fact that one variable has a direct effect on the other. This misconception implies that researchers only memorize simple relationship statements such as correlation ≠ cause (Huck, 2008).
Statistical significance results when a significant number of cases are evaluated leading to the establishment of a positive relationship between the variables (Ellis, 2010).Furthermore, Ellis (2010) opines that there exists a direct relationship between a study sample size and a high likelihood of having the null hypothesis rejected. When a larger sample size is selected, the probability of ascertaining a statistically significant result will be high notwithstanding a high possibility that the null hypothesis (H0) could be closer to being proven to be true . According to Valentine and Cooper (2003), a statistical significance test is used to confound two critical types of information namely, a) the degree of the expected effect and b) the sample size. Subsequently, statistical significance falls short of depicting the magnitude or size of the variable relationship effect since it does not illustrate practical significance also referred to as the relative impact.
SECTION 2: CORRELATION ANALYSIS
Since statistical significance is not linked to practical significance, it is unsuitable for use as a standalon...