Chapter Text
The model explains 60.6% of the variance in staring times (p < 0.001), suggesting a relatively strong explanatory power. Statistically significant predictors include the number of killings in the previous week, whether a dinner party was held, the sensible level of happiness, and the amount of alcohol offered to the researcher.
Regarding the traits of the Ripper, dinner parties are positively associated with staring times (β = 0.878, p = 0.024), while an increase in prior killings is negatively associated with staring (β = - 0.26, p = 0.036). Indicators of manipulation, including tearing, confession, and verbal manipulation do not show a statistically significant effect, meaning they do not interfere with the frequency of staring. Both predictors of Dr. Lecter’s positive mood, happiness and an increase in alcohol offered, are positively associated with times of staring. This suggests that Dr. Lecter will stare more when he is in a good mood.
The regression analysis indicates that Hypothesis 1 is partially supported. Dr. Lecter is indeed more likely to stare when a dinner party has been held. However, contrary to the hypothesis, the number of previous killings is negatively associated with staring times: the more killings he commits, the less frequently he stares. This may indicate that he is aware of the researcher's disapproval of increased killings and becomes more evasive and cautious during subsequent meetings. Alternatively, the negative association might reflect the researcher’s inability to detect his inappropriate behavior. The positive association between dinner parties and staring suggests that this behavior becomes more frequent when he is socially active. Notably, Dr. Lecter consistently attempts to invite the researcher to his dinner parties whenever possible. More data is needed to clarify the causal sequence of these actions. Given the mixed results, there is no strong evidence to support that staring is positively associated with traits of the Ripper.
Hypothesis 2 is entirely rejected. The level of manipulative acts does not significantly predict staring times. A separate correlation test shows that verbal manipulation is moderately correlated with confession, but there is no correlation between verbal manipulation and tearing. These results suggest that Dr. Lecter likely maintains a consistently high level of manipulation across all conversations, regardless of topic. Each confession appears to be meticulously planned, and he indeed cries frequently, even when there is seemingly no reason, perhaps simply to look magnificent.
Hypothesis 3 is strongly supported. Both indicators of Dr. Lecter’s good mood are positively associated with staring times. Notably, happiness exhibits the highest level of statistical significance, meaning it is almost certain that an increase in Dr. Lecter’s subjective happiness leads to more frequent staring. Additionally, increased alcohol offerings during sessions may lead to more physical movement, creating greater opportunities for staring. At the same time, Dr. Lecter appears to be less cautious after alcohol consumption, which may also contribute to a higher number of recorded instances by the researcher.
