Question

When we wish to make an inference about a population, we typically take a smaller sample that we can make generalizations from. Let's say that we wish to estimate the average height of all college students in the USA. We do so by asking every woman in CDS 101 to fill out an anonymous online form where they can report their height. What problems might prevent our generalization from being an accurate estimate of the average height of the entire population?
i. By only including women, we are not getting an accurate estimate of average height for our population of interest; therefore, we need to measure the men as well.ii. By looking only at students in CDS 101 at George Mason, we are taking a convenience sample rather than randomly sampling women from the whole population (all college students at all universities).iii. It is impossible to get an accurate estimate of average height without taking a census of all college students in the USA.iv. Some students may not not respond to the survey.

Answer 1

Answer:

i. By only including women, we are not getting an accurate estimate of average height for our population of interest; therefore, we need to measure the men as well.

ii. By looking only at students in CDS 101 at George Mason, we are taking a convenience sample rather than randomly sampling women from the whole population (all college students at all universities).

Step-by-step explanation:

Making inferences or generalizing sample statistics about a population can only attain a high level of accuracy by using a completely randomized survey or experimental design. This way the data will be free of bias as much as possible. In the scenario described above, the population of students is composed of both sexes 'Male and Female'. Therefore drawing a sample of only women to make generalizing about the population using such sample Data will be faulty.

Similary, by choosing the sample data from only one location, this means all others who should also have an equal chance of being chosen as part of the sample have been completely sidelined. This is another source of bias.

Hence. Theae two problems will prevent our generalization of the population from the sample from being accurate