Statistcs on provided data from mysql dump

We have an app collecting data on how long people take to learn morse code using different techniques ([login to view URL]): - With a auditory mnemonic cue ("A-part") - With visual mnemonics cue (Picture of A-Part) - With both - With just a morse sound We want to know questions such as: - How long does it take to learn morse code? - Using different techniques - how long does it take to learn? Whats the most effective technique? (and rank the different techniques) - Which letters cause the most problems? - How many people give up? can we infer when & why? Looking for someone to analyse this csv data. Note the user aggregate data tries to summarise data per user from the progress log data. The progress log data is the raw data sent every n seconds of playing. What I need is : - some stats / maths to calculate the above questions - but also some scripts that we can re-run on any new updated data Preferably Python based. Maybe in colab or jupyter notebook Some details on the data - two files. User aggregate data - which summarises per user data from progress log - the app pings the database every 30 seconds in play from each user - see the json field in progress dump. It’s a score per letter. 1 point for a correct letter. -2 for incorrect. 4 is learnt. So if a letter has 3 this could be 3 corrects or 100 incorrects and 49 corrects. (But unlikely. The point is it’s a little tricky to infer in detail from this) - time is all in seconds - if a user changes their settings they get a new user Id.