Question

Variables of interest: Levels of perceived power in authority and the perceived treatment of subordinates

Hypothesis: I hypothesize that there is a correlation in perceived poor treatment of subordinates and the increase of perceived power in superiors.

nteresting finding as it relates to my hypothesis: This 2018 study by Mead, et. al, was derived from Mead’s intrigue in why people misuse their power, coincidently. The study was an experiment that examines the manipulated hierarchical positions and testosterone levels. It was interesting that misuse of power only comes from people who have high testosterone. They found that power doesn’t corrupt everyone, only those who want the most.

Provide a meaningful and substantive analysis of hypothesis and on the application of their chosen empirical report. It should be at least five sentences.

Answer 1

Firstly note that the null hypothesis cannot be "there is a correlation between say X and Y". The null hypothesis is generally assumed to be a closed set or having disjoint support than that of the alternate. So, here we should have "there is no correlation between perceived poor treatment of subordinate and increase of perceived power in superiors" as the null hypothesis and "there is non zero correlation between perceived poor treatment of subordinate and increase of perceived power in superiors" as the alternate hypothesis. Considering the above as the null and alternate the following analysis has been performed.

According to the Mead study, higher testosterone level causes exploitation of one's power. These people crave for being at the top and often gets those positions through dominance. Thus, there exists a correlation between high testosterone level and being in leadership positions. This dominating nature leads to misbehaviour, ill-treatment of one's subordinates. Thus the variable "high testosterone level" directly influences both the concerned variables in the hypothesis (viz, perceived poor treatment of subordinate and increase of perceived power in superiors). Hence, it can be safely categorized as a confounding variable in this study. But, we know that if A is correlated with B and B is correlated with C then A and C can have no correlation. ( Take 100 tosses of a fair coin. A is the event of no. of heads in first 50 tosses, B is the event of total no. of heads in 100 tosses and C is the event of the no. of heads in last 50 tosses. Clearly, A and B are correlated, so is C and B. But A and C are not correlated) Hence, we cannot conclude about any correlation between perceived poor treatment of subordinate and increase of perceived power in superiors. So, due to the lack of evidence, we cannot reject the null hypothesis.