Experiment design for investigation of music semantic groupings

Experiment design for analysis of music descriptions

Instructions

Note - To test the experiment on your computer:

• Navigate to the project root directory in the terminal
• type python -m http.server
• Navigate to localhost:8000/test.html in web browser

When finished testing:

• type Ctrl-c to quit the python server.

Contributors

Catherine Deng, Maya Flannery

Hypothesis

Variables: Stimulus (4 * 14); Word lists: (14 X 38 X ??)

H_0: When a participant hears a stimulus, they select a random number of random words from each list. H_1: When a participant hears a stimulus, they select few random words from each list. H_1: When a participant hears a stimulus, they select few related words from each list.

We can hypothesize the number of genres selected (1 is most likely, 2, 3+ less common)

• use proportion? 10% of available words?

Instructions

Jupyter Lab

Using miniconda prompt:

1. Create environment (if first time running analysis)

> conda env create musicPython

2. Activate the environment

> conda activate musicPython

3. Install the required packages (if first time runnign analysis)

> conda install pandas matplotlib numpy nodejs nltk jupyterlab

4. Start jupyter lab

> jupyter lab

5. When finished, close jupyter lab, then in prompt:

> Ctl+C

6. Then type

> deactivate