“Dependent”, “Independent”, “Single and Double Blind” and “Placebo Effects”, Essay Example

Dependent Variable can be considered as the response which is measured in an effective manner (Harrington, 2009). Independent Variable, on the other side, is the value or the response manipulated by the researcher (Motulsky, 2010). Whilst the Independent Variable is the presumed cause and antecedent, the dependent one is the presumed effect or consequent. Within the chosen “Ashwini’ Hair Oil and Shampoo advertisement, the independent variable is the amount of shampoo to be used at every instance (which is controllable and manipulated by the user), whereas the dependent variable is effectiveness of Oil and Shampoo in preventing dandruff and offering scalp protection (Mitchell et al., 2009). In the field of non-experiemental research, Independent Variable possesses certain effect on a dependent one. Independent variables within the randomized experiments possess more than two levels, which define the experimental conditions. Randomization of subjects in any psychological or experimental research helps in offering a chance of ending up the condition with a higher weight of assignment. Single and Double Blind Studies for the selected advertisement, enable in exploring the uses in an in-depth manner. In the Single Blind studies, the subjects fail in knowing whether they are receiving an experimental treatment with Shampoo or Oil or a Placebo. Double Blind Studies, on the other side, involve no prior awareness to researchers and subjects on the treatment they have received-till the study is completed. The use of Double Blind studies in the current case, enable in decreasing the subject bias possessed by the researcher.

Placebo effect, is the measurable and observable effects or felt improvement in health and is not attributable to any form of treatment (Stevens, 2002). In the current case, the running of placebo-controlled and double blind studies could help in ascertaining the significance of selected product.

What do we mean by the statement, “Correlation is not causation”? Give examples for both.

“Correlation does not signify causation” is one amongst the most common phrase utilized in fields of Science (Psychology) and Statistics. This phrase signifies the fact that, correlation between two different variables may not automatically imply that one causes the other. Correlation, a remarkably simple concept, enable in establishing linear relationships between two random variables (Motulsky, 2010). Although, correlation is much needed for linear causation, in the non-existence of any third and countervailing causative variable, it illustrates possible causes and areas for further exploration. In theory, these are easy to differentiate-in a way an action or occurrence can cause another (such as asbestos exposure can cause Lung Cancer) or it can be related to another (like Asbestos exposure lead to other chronic respiratory disorders). In general, if a single action causes another, they appear to be most certainly correlated (Harrington, 2009). Unfortunately, our illustration may contribute us to astray when it comes differentiating between causality and correlation. For instance, having breakfast has longer been correlated with achievement of success in elementary school children. As a reason, it is easy to conclude that, eating breakfast can cause students in becoming better learners. However, those who do not have their breakfast appear to be most absent or tardy-and it is the absenteeism which is playing a crucial role in exhibition of their poor performance (Mitchell et al., 2009). Following the retesting of breakfast theory, researchers have found that, independent of other implications, breakfast only assists undemourished children in performing better. Numerous studies, which were actually designed to test a correlation offered suggestive reasons. In some cases, the correlated occurrences can be due to a common cause. This can be exemplified by the fact of red hair in correlation with blue eyes, emerges from a usual genetic specification. Also, correlation is usually witnessed when there is causality behind it. It is a well-established fact that, cigarette smoking not only leads to Lung Cancer, but also causes it. But, making an effort to establish such facts necessitates ruling out of possibilities- that smokers are likely to persist in urban regions, which are highly polluted. Despite of obvious facts that, one action causes another, more research is still needed to make numerous convincing arguments (Stevens, 2002).

What do we mean by the statement that there is statistical significant difference between two groups?

In the field of Statistics and Psychology, a statement called, “Statistically Significant” is used if it is likelihood of occurrence by chance is low. The word “Significant” in Statistics does not mean to be important or meaningful (Ziliak and McCloskey, 2008). This is usually represented through a p-value, which is the probability of random chance in explaining the result. In general, a p-value of 5% or lower is considered to be statistically significant. If there is, at most, around five percent of chance of occurrence between two events by coincidence, one can legitimately infer that, events have occurred together. The results in such cases are known to be statistically significant and events to be correlated (Motulsky, 2010). If the chance of random occurrence is more than five percent, the possibility of events to be happening together, solely by chance is much high to dismiss. Though, this five percent line is considered to be arbitrary, it has become standard in the field of biomedical research. Statistical significance, an extremely important parameter, is the golden measuring stick for data evaluation. Results, which are statistically significant, possess more weight in the scientific community (Stevens, 2002). Likewise, failure to demonstrate a conjectured correlation to be statistically significant can still offer good chances exploring the existence of real association. Statistical significance, therefore, must not be dismissed as lacking certainty. One can be 95% certain that a statistically significant result is true. Nonetheless, a small possibility that the result can be due to chance, which is why researchers keep their eyes open for other studies.

References

Harrington, A. (2009). The Placebo Effect: An Interdisciplinary Exploration, United States of America: Palgrave Macmillan Publishers.

Mitchell, A., Jolley, M and O’Shea, P. (2009). Writing for Psychology, London: Routledge Publishers.

Motulsky, H. (2010). Intuitive Biostatistics: A Non-Mathematical Guide to Statistics, United States of America: SAGE Publications.

Stevens, J. (2002). Applied Multivariate Statistics for the Social Sciences, United States of America: Tata McGraw Publishers.

Ziliak, T., McCloskey, N. (2008). The Cult of Statistical Significance: How Standard Errors Costs, London: Heinemann Publishers.