Exploring our repair data - what do you want to know?

data

#1

Together we’ve seen over 9,500 devices at more than 700 repair events. Thanks to our collective effort logging all that data, we’ve built up a rich database of the repairs we’ve attempted and types of devices we’ve seen.

But what stories can we tell with all this data?

We’ve always talked about wanting to use this data in interesting ways to help us all run better events and strengthen our case for the right to repair. But as a small team, we’ve sometimes struggled to make the most of all the data, which feels like a missed opportunity.

So, we’d love your help to explore the database and we’ve been working on a way to grant folks easy access to the data and powerful query tools with a platform called Metabase.
E.g.:

The data can be filtered in various ways - so if you’ve ever had any burning questions specifically about your own group’s repair outcomes, we can drill-down to those too :mag:

We have three questions for you:

:male_detective: What questions would you ask?
:woman_technologist: Do you want to dive into the data itself?
:man_artist: How can we tell interesting, data-driven stories?


#2

To whet your appetite, here’s a taste of the kinds of data we can pull out:

Repair success rate by category

Category Percentage Repaired Devices Repaired Total Devices
Paper shredder 75.8 % 25 33
Lamp 71.2 % 334 469
Toy 68.7 % 123 179
Headphones 65.3 % 164 251
Musical instrument 61.6 % 45 73
Laptop large 60.3 % 38 63


Number of devices seen by category


Repair outcomes


#3

What are the least repairable items, so we know what sections of the wiki need to be improved.


#4

Re: stories. If we can tell stories using the data and connecting that to everyday life, that might make the stories more powerful for the reader. Something like ‘1,500 laptops were reported repaired that would otherwise likely have been thrown away. It saved £xxxx in the purchase of new equipment, enough emissions to run a train around the world x times, etc.’. I hope the point is coming across :thinking:


#5

Love this! We can go deeper than what you suggest, like xx% of laptop repairs require spare parts. xx% of laptops were unrepairable by design, etc.

We have talented designer and dataviz friends who could help with this sort of thing. Would be amazing if volunteers themselves got stuck in, working through what equivalences might be appropriate/interesting, thinking about how they might be presented visually, etc.


#6

Hi, I’d like to look at what factors affect success rate, eg is it mix of products or may be how we define fixed. What could we do to increase success rate? Yes I’d be interested to dive into the data but not sure how much time I’ll have available.


#7

Knowing the %age repaired of each category could be turned into a powerful message when combined with the relative complexity of repair eg with paper shredders we’re really saying “Have a go, it’s usually a simple mechanical problem like a paper jam, there’s a good chance you can fix it yourself and our data supports that”. Ifixit provides a rating of complexity which could be used for this, or our own anecdotal evidence.


#8

Hi everyone, as a next step, we’re planning an event on 2nd March to explore our data on computers we fixed and those we couldn’t fix - I hope some of you can make it, all details here: