Question

Researchers conducted an observational study of automobile drivers and their cell phone use by manually observing traffic in heavily congested areas. They would record the amount of time they observed individual drivers on the phone while operating their vehicle in stop and go traffic and the number of accidents on a particular section of highway over the course of a month.

The accompanying graph depict the data they recorded, each point represents a 1 mile section of heavily congested roads.

Number of car accidents 1 10 2 3 4 5 6 į Š Average length of phone calls (minutes)

From the data, which of these is a potentially valid conclusion?

Car accidents and cell-phone use have no relationship.

The increasing rate of car accidents causes increased phone-call length.

A conclusion cannot be made; the data ignores the omitted variable of which cell-phone provider each person uses.

Cell-phone use while driving increases the rate of accidents.

b. If the xx and yy variables were swapped, what would the new slope be?

undefined

zero

negative

positive

Answer 1

a) Cell-phone use while driving increases the rate of accidents.

The scenario may include other variables also but this graph is specifically focused on driving through a congested area and talking on the phone while driving. The graph shows a clear positive correlation and that is quite linear in nature.
So option A and C are rejected. Option B is completely opposite in nature. It is not that the increased rate of accidents causes the increased length of the phone call but increased phone call length results in the increased rate of accidents.

b) Positive
The general practice is to plot independent variables on the X-axis and dependent variables on Y-axis. In this graph, the correlation between both variables is positive and almost linear. So if the variables were swapped, it will still have a positive correlation and linear structure. The slop will be positive for that graph.