Quality news or yellow journalism? These are the criteria

Every day, editors' mailboxes bulge with press releases about soundings and other market research. Often, these articles come from PR agencies, sometimes they contain 'real' news. How can one distinguish good research from bad? 

Is there research justification?

If researchers refuse to explain how they have obtained certain results and information, then do not publish their article.

Who has commissioned the article?

If the client could benefit from a certain outcome, then take that outcome (and the article as a whole) with a pinch of salt.

What is the target group?

If the research is aimed at a certain target group, the results will only apply to that same group.

How have the participants been selected?

In order to carry out a reliable research, a random sample of research participants must be selected. When respondents are able to participate voluntarily, representativeness cannot be guaranteed.

What is the sample size?

When few people are questioned, small differences are likely to be based on pure coincidence. If the sample size is 1000, then '50 percent' can be anything between 47 and 53 percent and will a difference of a few percents not mean that much.

What is the response rate?

If the nonresponse rate is 50 percent, the results will probably be very unreliable. The group that refuses to cooperate often holds different opinions than the group that does cooperate.

How is the list of questions drawn up?

If (a part of) your respondents consider the questions as leading, unclear or confusing, or if the order and introductions of the questions are leading, your research will be worthless, even if all other conditions have been fulfilled. If people did not have an opinion about a certain issue either, your research will be rather futile.

Source: Villamedia, ©Jop De Vrieze