Flight, Freeze, Or Fight
Though initially they may seem inefficient and somewhat counter productive, the freeze/fight/flight response to dangerous or threatening encounters that humans and other animals exhibit is actually a very useful way of dealing with many situations. Fighting or fleeing are really the only two ways to deal with a fatally threatening physical force; resisting or escaping are the only ways to remain alive. The initial "freezing" allows for a heightened perception of one's surroundings, which enables someone to determine the actual level of danger and, hopefully, the appropriate response to the situation (Bracha et al. 2004). In addition, this "freezing" has been shown to be an effective method of prey protecting itself against most mammalian predators, who tend to visually distinguish prey based on movement rather than color (Bracha et al. 2004). Remaining still renders one effectively invisible -- at least by sight -- to many predators, offering an extra level of protection. This also strongly suggests an evolutionary basis for the phenomenon, as man has not been a true prey animal since ong before the dawn of history. That, combined with the fact that many other animals lower on the evolutionary tree exhibit the same behavior, makes the evolutionary origin the most plausible.
Statistics
Real World Issue: Drunk driving and accident rates.
Independent variable: alcohol consumption.
Dependent Variable: accident rate.
Independent variable levels: six (none, one drink, two, three, four, more than four)
Null hypothesis: Drinking does not affect accident rates.
Alternative: Drinking increases accident rate.
Effect size would provide an indication of the actual correlative value between the variables, rather than simply commentary on their probable relationship.
Confidence intervals and means would provide easy summary material for the reader to determine the strength of the correlative relationship.
A post hoc test might implemented if accident rates were also seen to be significantly tied to age while drinking. This would be necessary to account for factors that could skew the data.
In an imaginary sample of 100 drivers who had at least one drink before driving, 34 were involved in accidents. This suggests a high correlation of drinking to accident rates when compared with an imaginary control group of 100 drivers that did not drink and drive.
References
Bracha et al. (2004). "Letter: Does 'fight or flight' need updating?" Psychosomatics 45, pp. 448-9.
Colorado State University. "Overview: Introduction to statistics." Accessed 30 July 2009. http://writing.colostate.edu/guides/research/stats/
