Doing Survey Research by Peter M. Nardi

Doing Survey Research by Peter M. Nardi

Author:Peter M. Nardi [Nardi, Peter M.]
Language: eng
Format: epub
Tags: Social Science, Sociology, General
ISBN: 9781317260974
Google: ZyzvCgAAQBAJ
Publisher: Routledge
Published: 2015-11-17T03:51:23+00:00


When we can specify the (nonzero) probability of each element in the population being selected for membership in the sample, we have what is called a probability sample. The most common types include

Simple random sampling

Stratified random sampling

Systematic random sampling

Cluster or multistage sampling.

Simple Random Sampling

Don’t be fooled by the word “random,” which seems to crop up regularly in everyday slang. It is a mistaken use of the word when someone tells you she stood on a street corner and “randomly” gave out questionnaires or walked up to people “randomly” in the mall or asked some “random” person to participate in a study. The researcher may have aimlessly wandered around finding people, but that doesn’t mean every person chosen was randomly selected. Did every shopper have an equal chance of being designated a respondent? Not likely, because some people were working, others shopped earlier in the day, and still others were hanging out at the other end of the mall.

Therefore, in order to achieve a true simple random sample, you must be able to provide a complete list of all possible units in the population from which to choose a sample. Whether you are randomly selecting ads in a magazine, students on a campus, or clients of a social service agency, somehow you must first get a complete and accurate set of elements according to the criteria you decide. For example, if you want to generate a sample of employees where you work, then you must first decide what an employee is (full-time, has worked at least six months, is not on maternity leave, etc.) and be able to get a complete list of all who fit the criteria. This becomes the sampling frame from which a sample can be chosen.

All units in the sampling frame may be identified with a number, either computer generated or done by hand, in order to use random sampling techniques. Each ad in a magazine could be assigned a number, every student already has an ID or mailbox number, each employee has a worker or social security number. Using either a table of random numbers (found in many statistics and methods books) or computer-generated ones (such as www.random.org/integers), the units of analysis are chosen. Or their names can be written on pieces of paper, placed in a box, mixed well, and then, like a lottery, picked out at random until the desired number of respondents is selected. Some computer programs can also generate random samples just using the names, so there is no need to first assign numbers to each element in the sampling frame.

Telephone surveys employ probability sampling through random digit dialing techniques in which machines generate phone numbers within various area codes and then dial the numbers. Even those with unlisted numbers can be selected using this method. However, people without phones or who use only mobile phones are not part of the sampling frame, and this possibility could bias the resulting sample by underrepresenting the poorest people. See Box 5.3 for a description of this technique as used by a national polling company.



Download



Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.