For example, a company from Magnitogorsk produces household goods. To develop a product line, it wants to study people’s attitudes to holidays and explore traditions of decorating an apartment for different dates. That is, in this case, your target audience is the entire population of the city. It is called the general population.
Let’s say you know from the census on the Rosstat website that one third of the city’s population are men and two thirds are women. In order for the survey results to be accurate, you need to develop a sample that will respect these proportions.
The accuracy of the study also depends on the number of respondents
It is not enough to ask questions to one man and two whatsapp number list women. You need to talk to such a number of people who will ensure the correctness of the results.
Important point: the sample is modeled according to the parameters that significantly influence the participants’ answers. Most often, these are gender, age, income level, place of residence. Sometimes specific factors are used, such as profession, use of the one of the main problems associate Internet, availability of a car, or special expense items.
To solve the problem of studying holidays, it is enough to divide respondents by gender and age. But if the company sells expensive attributes for the home, then it is worth using income as well. Then you will receive information about the opinion of each segment of the audience – economy, middle and VIP.
When a representative sample is needed
Representative sampling is used in marketing research . There are two types of it – qualitative and quantitative.
In qualitative research
Qualitative research is focus groups and in-depth interviews. They mobile lead are needed to find new ideas and hypotheses, answer the questions “Which?”, “Why?”, “Why?” and so on.
Their main purpose is to get an idea of the diversity of opinions of the target audience , rather than to form an accurate picture of the distribution of these opinions. Therefore, the sample for qualitative research can be representative, but small in terms of the number of respondents.
