I need help to refine my Online Javascript/HTML Web Experiment for participant testing

I have an online Javascript, HTML and CSS project. It is an online game and experiment with two parts. In the first part, the player is tasked with using instructions from a generated language to place different coloured blocks in a specific order. In the second part, the player then looks at each task, and writes an instruction on how to complete it.

This is the link to the current deployment: [login to view URL] .

The project code is available at: [login to view URL] .

In order to start testing with participants, I need changes made to refine the project. There are three areas which need upgraded.

Language Evolution

Currently, words of the artificial language are generated by choosing a batch of 5 random tokens from /json/[login to view URL] , and then assigning these tokens to English words based on /json/[login to view URL] .

This will create the first generation of the language.

In part two of the experiment, the player looks at each task, and uses the available tokens to write an instruction for how to complete it.

I need this user input to affect the instructions that the next player sees, effectively evolving the language over time.

Specifically, what I need is for each instruction from the most recent player to be taken, and then have every unique ‘token’ in that instruction replaced with a new one, which is randomly selected from a new batch of 5 random tokens.

For example, if the previous participant’s instruction read: ta wuta toto

Then it could be randomly replaced with: ke boke wewe

This replacement will become the new language, so when the next participant tries part 1 of the experiment, the instructions given to them should be from what the previous participant wrote, with token replacement as described above.

User Experience

In general, I need the experiment to become more user friendly and encouraging.

For part 2 I need the process to be faster since it currently takes too long to complete.

In particular, I would like the tasks in part 2 to be cut down to remove tasks 1,2,4,6 and 7.

To stop participants from overthinking, I would like for the instructions at the beginning to be edited and tell the players:

Please do not write down the meanings of words, and it is ok to make mistakes, simply put the answer that first comes to mind. If you cannot find the correct solution, please make a few attempts and then it will move onto the next task.

I would like a visible time limit of 30 seconds per task in part 2.

The keyboard in part 2 should be changed so that it only shows the 5 tokens currently in use.

Reliability

At the moment, there is a reliability issue in how the participant’s data is sent to the web server.

Currently, after a participant completes the final survey of the experiment, their results data in the form of a json called game_json is received at the web server called somata, the data of which can be viewed at [login to view URL] . It is very important that after all of the changes, the results of a completed experiment run will still appear at this link.

One requirement is that I need the sending of the data to the server to become more reliable. This could be done by sending the game data incrementally throughout the process, rather than right at the end after the survey is completed. This is in case a participant exits the game too early or experiences a crash.

Deadline

I will need the project to be ready by the 17th of March 2021

JavaScript HTML CSS Anglická gramatika English (UK) Translator

Identifikační číslo projektu: #29443786

O projektu