Question

A random sample of adults in a large US city responded to a series of questions about their use of public transportation. The results of the survey showed that 520 of the 838 adults in the sample (62%) have used public transportation in the past week. How might the results of the survey be biased in obtaining an estimate of all US residents who use public transportation?

A. Because not all residents in this US city were asked to participate, the survey may underestimate usage of public transportation.
B. Because the sample includes only residents in one US city, the survey may overestimate the proportion of US adults who use public transportation.
C. Because those surveyed may choose not to answer, the survey may underestimate the proportion of US residents who use public transportation.
D. Because the survey includes a random sample of residents in one US city, the survey should provide an accurate estimate of the proportion of US adults who use public transportation.

Answer 1

The survey may overstate the proportion of US adults using public transportation as it solely includes residents from one city, limiting its generalizability to the entire population. The correct answer is B.

Step-by-step explanation:

In statistical terms, the bias in this survey is due to a lack of geographical diversity in the sample. Option B correctly identifies that the survey might overestimate the proportion of US adults who use public transportation because it only includes residents from one city. Public transportation habits can significantly vary across cities and regions due to differences in infrastructure, urban planning, and cultural preferences.

For a more accurate estimate of the proportion of US adults using public transportation, a nationally representative sample is essential. Including residents from various cities and regions would provide a more comprehensive view of transportation habits. The bias in this survey arises from the limited scope of the sample, making it challenging to generalize the findings to the entire US population.

While a random sample is a good practice in survey methodology, the potential overestimation in this case is a result of a specific and non-representative geographic focus. It's crucial to consider regional variations when drawing conclusions about broader populations, and in this instance, the bias is introduced due to the survey's limited coverage of only one US city.