Experimental Setup - UserModels2223/group-project-m-f-h GitHub Wiki

Experimental setup

In order to test if we can successfully create a user model that is capable of determining whether or not a user learns better using purely visual stimuli or if they would benefit from audio + visual stimuli, we run the following experiment:

Our test subjects will study Portuguese using our adaptation of SlimStampen. As in the default SlimStampen, the user gets prompted with a word written in Portuguese (visual stimuli), and will have to type out the English translation. The first time each word is introduced the user is given both the Portuguese prompt as well as the translation. And in the case that the system decides to present a word as audio + visual stimuli, the Portuguese word is shown as usual, and at the same time the test subject will hear the Portuguese pronunciation of the word via headphones. We chose to use headphones as it will better isolate the user in the experiment to minimize possible distractions.

While they are repeating words and learning new words, the system keeps track of what condition elicits the best performance from the particular user, and after 5 minutes of studying it will start increasing the chance of one of the two conditions based on this. Then for the remaining 5 minutes the user will continue to use the system which will continue to update the chance for either condition based on performance of the user.

An example of the flashcard system can be seen here:

Here the word sangue and its translation are given as the word is presented for the first time. Afterwards, is the word sangue is presented again, the translation will not be given. However, the audio file will still play.

Testing

We plan to gather around ten participants to perform our experiment. They will be around the age of 20-30 and mostly Dutch.

Participants will be placed in a quiet room behind a laptop running our system through OpenSesame and will wear a headset to hear the audio files. The total duration of the experiment will be around ten minutes.

Before they use SlimStampen, we require the participants to fill in a questionnaire in which we ask some questions about them such as what languages they are proficient in or familiar with to see if this might impact the results. The questionnaire can be found below.

Google form:

https://docs.google.com/forms/d/e/1FAIpQLSecXP6ctqvDo0f-10vZS61Z15v01hR34tUWBtgwxj0B4CQNDQ/viewform