Welcome to the dB-SPLab testing page!

Here you can have lots of fun doing one of our online experiments. Check the list below to see what's poppin' at the moment. But before you head on to a world of adventure and wonder, please read our foreword carefully. This is the small print part of our website, but we printed it big so it's easier to read.


We are the Speech Perception research group at the Department of Otorhinolaryngology of the University of Groningen, located at the University Medical Center Groningen. You can find information about our research on our other website. We run experiments on (human) volunteers to better understand how people — and especially those with impaired hearing — perceive speech in difficult communication situations, for example, when there is some noise in the background.

All our experiments have been approved by the Medical Ethics Committee of the University Medical Center Groningen (METc UMCG) to ensure we follow ethical guidelines and properly protect personal information. In all experiments, you will answer some questions, or some sounds will be played on your computer and you need to click somewhere on the screen to give your answer.

We take data protection very seriously. We will never share your personal data with anyone outside of UMCG. We also apply state-of-the-art encryption methods to make sure your data is safe with us. Your data is separated in two parts: (1) identification data; and (2) experimental data.

Identification data

We need to collect identification data as proof that you gave us informed consent before taking part in the experiment. This is a corner-stone of ethics principles when working with human volunteers.

We also need a personal identification token to make sure we did not inadvertently test you twice on the same experiment. For that purpose, we ask you to enter your email address as initial identification. We derive a unique key from this email address, that allows us to know if we have seen you before. But it is not possible for us to reconstruct your email address from the key (you can find information about the algorithm we are using here). Here's an example:

jan.jansen@mail.com ===> f7cebc2613b9b641078ea0fb49e4f62e
f7cebc2613b9b641078ea0fb49e4f62e =X=> ???????????????????

You are, of course, free to use something else than your email address as identification, as long as you always use the same identification token.

Finally, to be able to identify you (or actually, your browser) throughout the experiment, we use a session mechanism that relies on cookies. These cookies are normally destroyed at the end of the session, or otherwise they expire a few hours after you visited our website. In any case, these cookies are not used to further track your behavior on the internet outside of the experimental session, and the identification data relative to them are never transmitted to third-parties.

Experiment data

The purpose of the experiment is of course to collect the experimental data, that is, the answers to our questionnaires and which buttons you are clicking in response to the sounds that are played. The experimental data is pseudonymized, which means it is disconnected from your identification data. While there is a connection table that does exist, in normal operation we do not access it. We believe in Open Science, so your pseudonymized experimental data will be shared with colleagues within and outside the UMCG. We will use it to present the results of our investigations at international conferences and to publish articles in international journals. However, your data will always be presented in the context of other data collected from other volunteers and cannot be traced back to you as an individual.


Here is the list of experiments we are running at the moment. Each experiment has a set of inclusion criteria that are listed on the right-hand side. To start an experiment, just follow the corresponding link:

➡︎Hoe klinkt een blije stem? Luisteren naar stemmen en emoties
  • Over 18.
  • Moedertaal Nederland.
  • Geen auditieve of spraak probleem.