Explore BrainMass

Sampling methods and level of measurements

1) American On Line ran a poll of its Internet subscribers and asked this question - Should the World Trade Center towers be rebuild?" among the 1,304.240 response : 768, 731 said Yes and 286, 756 answered NO and 248,753 said that they were "undecided". Being that this sample is extremely large can the responses be considered to be representative of the population of the United States?

2) Visa wants to conduct a survey of credit card usage among the full-time students who attend college. What is a good procedure for obtaining a sample of each type: random, systemic, convenience stratified, cluster.

3) a) the weights of people in a sample of elevator passengers
b) a movie critic's ratings of "must see; recommended; not recommend; don't even think about going"
c) a movie critic's classification of "drama, comedy, adventure"
d) Bob, who is different in many ways measures time in days, with 0 corresponding to his birth date. The day before his birth is -1 , the day after his birth is +1 and so on . Bob has converted the dates of major historical events to his numbering system. What is the level of measurement of these numbers.

IDENTIFY WHICH OF THE ABOVE IS Nominal, Ordinal, Interval, Ratio level of management.

Solution Preview

There are a number of limitations associated with study. The foremost of which is the lack representativeness of the sample. The sample selected is not random. Only subscribes of American Online can participate in this study. Randomness is the fundamental principle of any statistical study.

The Internet population has a slight bias towards younger, more computer literate users. The study limited the potential pool of respondents to those already involved in online activities and some members of the target population may have not participated because of their discomfort with on line interactions.

Thus in the above study sampling error is not controlled. It can be noted that the sampling error tend to increase with increasing sample size. Thus sample collected is not a representative of the population of the United States.


1. Hamilton, J. C. (1999). The ethics of conducting social science research on the Internet. The Chronicle of Higher Education, 46 (15), B6-7.
2. Stanton, J. M. (1998). An empirical assessment of data collection using the Internet. Personnel Psychology, 51 (3), 709-725.

2) Visa wants to conduct a survey of credit card usage among the full-time students who attend college. What is a good procedure for obtaining a sample of each type: random, systematic, convenience stratified, cluster?

Random sampling
A Simple ransom sampling procedure that assures that each element in the population has an equal chance of being selected. Let us assume you there are N students, and you wanted to select n of them for further study. You might put all their Names / id numbers / Visa Card Number in a drum and then pull n names out. This means that every student has a n / N chance of being selected using this method. You can also use some statistical package like SPSS to generate random numbers of id of students

Stratified sampling

Using stratified sampling, the population is divided into ...

Solution Summary

The solution gives the details random sampling, systematic sampling, convenience sampling, stratified sampling, cluster sampling and other sampling methods. Definition and examples of nominal, ordinal, interval and ratio level of measurement are also given.