Selecting a Sample

Many researchers feel that there is nothing more important than selecting a sample that accurately reflects the characteristics of the population they are interested in studying. Yet sample selection can sometimes be a risky business, with all kinds of questions needing to be answered before you can make any moves toward the sample selection process. Here is a list of factors to keep in mind:

• 1. Imagine yourself trying to find a suitable pool of candidates from which to select a sample, and multiply the number of other people trying to do the same thing in your community by 100. That is a small estimate of how many people in every university community are looking for a sample to include in their study. Where can you look? Try some of the following:
  • • Church and synagogue groups
  • • Boy and Girl Scouts
  • • Retirement homes and communities
  • • Preschools
  • • Singles clubs
  • • Special interest and hobby groups
  • • Fraternal organizations
• 2. Remember, you do not want to select any group that is organized for a particular reason if that reason is even remotely related to what you are studying. For example, you would not select members from the Elks Club for a study of loyalty or friendship or parents who send their kids to private schools for a survey on attitudes toward supporting public education, unless the selection of such samples is an important part of your sampling plan.
• 3. Approach candidates with a crystal clear idea of what you want to do, how you want to do it, and what they will get in return (a free workshop, the results of the study, or anything else you think might be of benefit to them).

Similar to the previous point, the population must match the characteristics of those groups you want to study. It might go without saying (but I’ll say it here anyway), but selecting a sample from a poorly identified population is the first major error in sample selection. If you want to study preschoolers, you cannot study first graders just because the two groups are close in age. The preschool and the first-grade experience differ substantially.The type of research you do will depend on the type and size of sample you need. For example, if you are doing case study descriptive research, which involves long, intense interviews and has limited generalizability (which is not one of the purposes of the method), you will need very few participants in your sample. If you are doing a group differences study, you will need at least 30 participants for each group.A highly reliable test will yield more accurate results than a homemade essay exam. The less reliable and valid your instruments, the larger the sample size that will be required to get an accurate picture of what you want.Consider the amount of financial resources at your disposal. The more money (and resources in general) you have, the more participants you can test. Remember, the larger the sample (up to a point) the better, because larger samples come closer to approximating the population of which they are a part.The number of variables you are studying and the number of groups you are using will affect the sample selection process. If you are simply looking at the difference in verbal skills between males and females, you can get away with 25–30 participants in each group. If you add age (5- and 10-year-olds) and socioeconomic status (high and low), you are up to six different possible combinations (such as 5-year-old girls of high socioeconomic status) and up to 6 × 30, or 180, subjects for an adequate sample size.

Data Collection an