You are on page 1of 1

The article chosen is from CNN describing Tesla car autopilot.

The
maker of Tesla cars claims that autopilot reduces driver workload and
improves road safety despite recent crashes. "Nonetheless, when used in
conjunction with driver oversight, the data is unequivocal that Autopilot
reduces driver workload and results in a statistically significant improvement
in safety." (McFarland, 2016)

McFarland, M. (2016, July 1). Tesla's autopilot probed by goverment after


crash kills driver. Retrieved August 1, 2016, from CNN Money:
http://money.cnn.com/2016/06/30/technology/tesla-autopilot-
death/index.html

The null hypothesis for the claim being examined in this case is that
Tesla autopilot has no effect on driver workload or overall safety. In other
words, the overall driving experience is not changed by autopilot in Tesla
cars. The alternative hypothesis for the claim being examined is that Tesla
autopilot reduces driver workload and overall safety. In other words, driver
experience is improved by the use of Tesla autopilot system.
The article claims that there was a significant improvement in the
safety of drivers when autopilot is used. This claim is the equivalent of
rejection of the null hypothesis. The risk of deciding to reject the null
hypothesis is that of a false positive or a type 1 error. A false positive in this
context would be an improved driver experience and safety when there in
reality is none. The result was determined to be within some margin of error
or low false positive rate because the result was statistically significant.
Here, this could be potentially very dangerous. Although the null hypothesis
is that there is no change in driver safety, another alternate fate is that the
overall safety experience is worsened by the use of the autopilot. The article
describes a car crash that was attributed to the use of the autopilot. This
could be an example of a false positive, where the autopilot actually has
poorer safety but is attributed to an increase safety. This would lead to a
higher incidence of accidents on the road and ultimately hurt those using the
autopilot and other drivers on the road.
In the context of this article, the difference between practical
significance and statistical significance are not discussed. However, we can
discuss the implication of the difference in context of this case. If the
statistical significance was determined based upon some score of safety,
then there is some numerical difference between usage of autopilot and non-
usage. If this difference is marginal but significant that have implications for
the actual usage of the service. If autopilot only improves safety marginally
on the roadway, it may not be worth the hassle and risk. This is especially
true if autopilot uses extra gas or has increases emissions.

You might also like