7. Logic of Sampling Jin-Wan Seo, Professor Dept. of Public Administration, University of Incheon

7. Logic of Sampling 7. Logic of Sampling 1  Sampling  Social researchers must select observations that will allow them to generalize to people and events not observed. Often this involves sampling, a selection of people to ob serve. The process of selecting observation is _____________. Introduction

7. Logic of Sampling 7. Logic of Sampling 1 Sampling in social research has developed hand in hand with political polling.  Famous poll conducted by the Literary Digest (1890~1938) in 1920  mail postcard to people in 6 states, asking them who they were planning to vote for in the presidential race between Warren Harding and James Cox.  names are selected for the poll from telephone directories and automobile registration lists  Based on the postcards sent back, the Digest correctly predicted that Harding would be elected.  In the election that followed, the Digest expanded the size of its poll and made correct predictions in 1924, 1928, and Brief History

7. Logic of Sampling 7. Logic of Sampling 1  In 1936, the Digest conducted its most ambitious poll.  Over two million people from ten million responded, giving the Republican contender, Alf Landon, a stunning 57 to 43 percent landslide over the incumbent, President Franklin Roosevelt.  Two weeks later, voters gave Roosevelt a second term in office by the largest landslide in history, with 61 percent of the vote. Landon won only 8 electoral votes to Roosevelt’s 523.  Why? Sampling problems – exclusion of poor people.  The poor voted predominantly for Roosevelt’s New Deal recovery program. The Digest’s poll may have correctly represented the voting intentions of telephone subscribers and automobile owners.  George Gallup correctly predicted that Roosevelt would beat Landon. His success in 1936 hinged on his use of something called quota sampling.  Gallup and his American Institute of Public Opinion used quota sampling to good effect in 1936, 1940, and Brief History

7. Logic of Sampling 7. Logic of Sampling 1  In 1948, failure of the Gallup  In 1948, Gallup and most political pollsters predicted that Governor Thomas Dewey of New York would win the election over the incumbent President Harry Truman.  Why? Sampling problems too – unrepresentativeness of his samples  Quota sampling – which had been effective in earlier years – was Gallup’s undoing in  This technique requires that the researcher know something about the total population – in this case, the population of voters.  For national political polls, such information came primarily census data. By 1948, however, World War II had produced a massive movement from the country to cities, radically changing the character of the U.S. population from what the 1940 census showed, and Gallup relied on 1940 census data.  City dwellers, moreover, tended to vote Democratic; hence the overrepresentation of rural votes in his poll had the effect of underestimating the number of Democratic votes. Brief History

7. Logic of Sampling 7. Logic of Sampling 1  Probability Sampling & Nonprobability Sampling  Sometimes you can and should select _____________ sampling using precise statistical techniques as the primary method of selecting large, representative samples for social science research including national political polls.  At the same time, probability sampling can be impossible or inappropriate in many research situation. In this case, __________________ sampling techniques are more appropriate. Two Types of Sampling Methods

7. Logic of Sampling 7. Logic of Sampling 1 1. Reliance on available sampling  Relying on available subjects, such as stopping people at a street corner or some other location, is an extremely risk sampling method.  This methods does not permit any control over the representativeness of a sample.  It is justified only if the researcher wants to study the characteristics of people passing the sampling point at specified times or if less risky sampling methods are not feasible.  Even when use of this method is justified on grounds of feasibility, researchers must exercise great caution in generalizing from their data. Also, they should alert readers to the risks associated with this method. Nonprobability Sampling

7. Logic of Sampling 7. Logic of Sampling 1 2. _______________ or judgmental sampling  Sometimes it’s appropriate to select a sample on the basis of knowledge of a population and the purpose of the study.  In the initial design of a questionnaire, for example, you might with to select the widest variety of respondents to test the broad applicability of questions. Although the study findings would not represent any meaningful population, the test run might effectively uncover any peculiar defects in your questionnaire. This situation would be considered a pretest, however, rather than a final study. Nonprobability Sampling

7. Logic of Sampling 7. Logic of Sampling 1 3. Snowball sampling  A form of ___________________  This procedure is appropriate when the members of a special population are difficult to locate, such as homeless individuals, migrant workers, or undocumented immigrants.  The researcher collects data on the few members of the target population, he or she can locate, then ask those individuals to provide the information needed to locate other members of that population whom they happen to know.  “Snow ball” refers to the process of accumulation as each located subject suggests other subjects. Nonprobability Sampling

7. Logic of Sampling 7. Logic of Sampling 1 4. Quota sampling  Quota sampling begins with a matrix, or table, describing the relevant characteristics of the target population.  Once you have created such a matrix and assigned a relative proportion to each cell in the matrix, you proceed to collect data from people having all the characteristics of a given cell. You then assign a weight to all the people in a given cell that is appropriate to their portion of the total population.  Inherent problems (differ from _____________ sampling)  the quota frame must be accurate and it is often difficult to get up- to-date information for this purpose  the selection of sample elements within a given cell may be biased even if its proportion of the population is accurately estimated. Nonprobability Sampling

7. Logic of Sampling 7. Logic of Sampling 1  Key  The key to probability sampling is _____________ selection.  Probability sampling methods provide an excellent way of selecting representative samples from large, known populations.  Problems  These methods counter the problems of conscious and unconscious sampling bias by giving each element in the population a known probability of selection.  The most carefully selected sample will never provide a perfect representation of the population from which it was selected. There will always be some degree of sampling error.  Probability sampling methods make it possible to estimate the amount of sampling error expected in a given sample by predicting the distribution of samples with respect to the target parameter. Probability Sampling

7. Logic of Sampling 7. Logic of Sampling 1 1. Simple random sampling  logically the most fundamental technique in probability sampling, but it is seldom used in practice. 2. Systematic sampling  involves the selection of every kth member from a sampling frame. This method is more practical than simple random sampling, and, with a few exceptions, it is functionally equivalent. Probability Sampling

7. Logic of Sampling 7. Logic of Sampling 1 3. _________________ sampling  Stratification, the process of grouping the members of a population into relatively homogeneous strata before sampling, improves the representativeness of a sample by reducing the degree of sampling error. 4. Multistage cluster sampling  a relatively complex sampling technique that is frequently used when a list of all the members of a population does not exist. Typically, researcher must balance the number of clusters and the size of each cluster to achieve a given sample size.  Stratification can be used to replace the sampling error involved in multistage cluster sampling. Probability Sampling