ultrafabber
Member
THIS IS THE CHART : https://i.imgur.com/Nt7ma2B.jpg
This is a follow-up of a chart I made a while ago https://preview.redd.it/ce6j48jfzj921.jpg?width=960&crop=smart&auto=webp&s=1617ad8f95de920491115f474b7b3218f205f2d6
The initial topic contained the data for a single day. That is, i went in and gathered the data on how many people were in each streak timeframe at a particular point in time (January 10th). Since it was a single day observation, there was a significant possibility that the data was not overall representative of a trend, even if the sample size is very large.
So I gathered the data over a period of 28 days and what you see in the chart are all the entries overlapped, which demonstrates an obvious trend.
I have the data for more than 90 days - it's up to 700 days, but i narrowed the chart to 90 days because it would've made the graph unreadable. You are more than welcome to play around with the data and come up with other statistics and visualizations. Filebin link that will expire in one month from now https://filebin.net/22oq0vyu3p94btb3
Limitations of the results:
1. the data is gathered from a very popular app but it's limited by how the app creator decided the progression cutoff points (3 days, 5 days, 7 days, 10 days etc). Obviously it would've been much better if we had data on every single day, not on irregular intervals but it's the best we have so far and the sample is huge.
2. the app creator told me that inactive users are removed after 7 days. Inactive means they haven't logged in the app for over 7 days.
3. the fact that inactive users are removed has upsides and downsides, it doesn't inflate success rates but at the cost of inflating failure.
4. i missed 4 days and i filled in the numbers by averaging the former and prior numbers.
This is a follow-up of a chart I made a while ago https://preview.redd.it/ce6j48jfzj921.jpg?width=960&crop=smart&auto=webp&s=1617ad8f95de920491115f474b7b3218f205f2d6
The initial topic contained the data for a single day. That is, i went in and gathered the data on how many people were in each streak timeframe at a particular point in time (January 10th). Since it was a single day observation, there was a significant possibility that the data was not overall representative of a trend, even if the sample size is very large.
So I gathered the data over a period of 28 days and what you see in the chart are all the entries overlapped, which demonstrates an obvious trend.
I have the data for more than 90 days - it's up to 700 days, but i narrowed the chart to 90 days because it would've made the graph unreadable. You are more than welcome to play around with the data and come up with other statistics and visualizations. Filebin link that will expire in one month from now https://filebin.net/22oq0vyu3p94btb3
Limitations of the results:
1. the data is gathered from a very popular app but it's limited by how the app creator decided the progression cutoff points (3 days, 5 days, 7 days, 10 days etc). Obviously it would've been much better if we had data on every single day, not on irregular intervals but it's the best we have so far and the sample is huge.
2. the app creator told me that inactive users are removed after 7 days. Inactive means they haven't logged in the app for over 7 days.
3. the fact that inactive users are removed has upsides and downsides, it doesn't inflate success rates but at the cost of inflating failure.
4. i missed 4 days and i filled in the numbers by averaging the former and prior numbers.