Why does reject sampling work

The selection bias in marketing

The selection bias has long been of great relevance not only for research. In business life, even in private everyday life, we also select information or already receive it selectively distorted data presents. Cognitive distortions also make a major contribution to the fact that we make mistakes when making a selection, which inevitably falsify a result.

The recurring sample biases clearly show that we not impartial but have to invest a lot of work in order to at least approach a status of impartiality. The following examples of selection bias show the far-reaching implications of sample bias.

In the first example, a General brand awareness survey of a healthy dietary supplement. If the survey is carried out in fitness studios, in health food stores or in organic supermarkets, the target groups of the products are asked. This can be useful. In any case, the results of the market research should be treated with caution, because the selection bias has already struck: The people who visit the gym, health food store or organic supermarket are usually more receptive to the effectiveness and usefulness of healthy products. It can therefore be assumed that brand awareness is higher in these groups of people - and that this was accordingly not measured neutrally.

The second example of selection bias shows the far-reaching consequences of not using true random selection. Economic researchers should be one Survey on the economy that is as representative as possible of all companies in the country. The data is selected on the basis of the commercial register and the corporations and trading companies entered there. The selection bias is even greater here than one might assume at first glance: the bias in the sample not only excludes small businesses, but also numerous successful freelancers (e.g. lawyers, doctors, architects), artists and Part-time occupation in all industries.

This example is obvious - experienced researchers will not make such a mistake. However, many smaller sample distortions can add up and thus greatly falsify even such an important number as the economic forecast for a country.