Polindex: Why good presentation matters - a lot

This article is provided to you free of charge. If you want to support our team, subscribe now.

With more than 1,500 respondents, around a year to process and analyse, and a 154-page presentation of the results, no one can doubt that many hours of work and even more resources have gone into the joint project of the Chaire de recherches en études parlementaires of the University of Luxembourg and the survey company ILRES. It could, indeed should, be the most revealing political survey of the year: a much-needed reference work for journalists, a decision-making aid for politicians (to whom the study was presented in the Chamber) and a treasure trove for all institutions and those interested in politics.

It is therefore all the more incomprehensible that its graphic presentation unfortunately makes it almost impossible to use it for more than the roughest guesstimations. The presentation is a colourful mixture of small careless mistakes and ambiguities that should actually have been improved in the first draft: Missing labels, incorrect legends, responses that are accompanied by terse interpretations of the results but – almost unbelievably – not the questions themselves, or presentations that could at best hint at possible correlations.

Yet the brains behind the study – none other than the university and ILRES – should know better. Both institutions regularly publish high-quality studies that present far more complex data and correlations in a comprehensible, often even attractive manner. This raises the question: What has happened here? Especially since the report of previous years had exactly the same problems? It seems as if they always stumble in the final stretch.

Let's take a look at some examples. If you want to read along, you can download the report here: PDF-LINK

Here, dear readers, is a screenshot of a simple table and a question: What exactly can we read from this bar chart (p.73)?