It is almost impossible and above all very expensive to interview all potential customers . The essential thing remains that the sample be representative of the population to be studied. Determining the sample size will need to take into account the budget, the degree of detail and the time available to carry out the survey. Obviously, the larger the sample, the more reliable the questionnaire will be, but the more data you have to analyze, the more time-consuming or resource-intensive administration can be. Sampling methods 1/ The quota method The method generally chosen is the quota method because of its simplicity. It is a question of taking into account the statistics first.
The Most Used Criteria Are Generally
Other criteria can be used. It is obvious that the more the number of criteria increases, the more the administration of the questionnaire will be difficult. Example on a population of 500 people: it will be necessary to multiply the percentage (in terms of distribution) by the desired number to know how many people are to be questioned. 2/ The route Religious Email List method This is a method where it is not a question of selecting the interviewer according to criteria such as age and sex but following an initial instruction. This involves, for example, questioning the inhabitants of a street with even numbers. 3/ Cluster sampling It is a question of segmenting a population in particular by geographical space.
Example the 20 Arrondissements of Paris
Then it is necessary to select certain clusters at random and to identify. All the individuals of the selected clusters in order to question them. If the information is not available on a cluster, then a new draw must be made. Sampling by strata This method requires an exhaustive list. A very good knowledge of the distribution of the population studied. This involves Ao Lists determining the number of individuals to interview. Per stratum (sex, age, etc.) and drawing lots per stratum Means of administering the questionnaire. The method to be selected depends on several factors: cost of the desired survey. Number of people to interviewe type of survey, type of information and desired quality.
