Population into groups relevant to our interests

Sex or age affects buying behaviour

Should not try to imitate them by using their slang or accents.R ather, we should use simple, easily understood language, defining any terms which the respondent seems not to understand. The most common method of selecting a few representative individual from a large population of potential respondents is through random sampling. In effect, we put the names of the entire population in a hat and then pull out the desired number of infalmants. This "drawing from a hat" technique can be accomplished in a variety of ways but with each method the principle remains the same: everyone in the population has an equal chance of being chosen, and the selections are made purely on the basis of chance.

Other sampling techniques are systematic, stratified, and quota sampling. When sampling systematically, we develop a system by which the sampling will be chosen. We might decide to select every twenty-fifth name in a telephone book, using the resultant group as our sample. Stratified sampling is somewhat more complicated. We divide the population into groups relevant to our interests and then randomly select respondents from those groups. If we are conducting a survey of buying habits, we might want to learn whether sex or age affects buying behaviour. Thus we would divide out population into males and females and young and old-and from the four groups select randomly representatives to be interviewed. The quota method involves selecting a sample which matches as closely as possible the characteristics of the entire population. If we still are interested in the effects of sex and age, and we discover that our population consists of 40 per cent old women, 20 per cent old men, we would select a sample of these same proportions. Although each of these sampling methods involves somewhat different processes, the end is the same: to reduce to a manageable