Citizen Science on the web is about more than just collecting data

We won the project, thanks in part to our extensive experience in online voting systems for ITV and the National Lottery People’s Projects.

After a hectic month or two, learning loads and getting our hands dirty — metaphorically speaking — The Big Compost Experiment launched.

Here are some of the things we learned about citizen science on the web...

Mark and Danielle were leading an investigation into the role and effectiveness of so-called biodegradable and compostable plastics; the kind that is often accompanied by eco-sounding messages like “I’m made out of potatoes” or “I’m a 100% compostable plastic spoon”.

The study had two clear stages.

Firstly, they wanted to know what people thought about biodegradable plastics. Did biodegradable packaging influence purchasing decisions? Was there a clear understanding in the population about what ‘biodegradable’ means? Were people actually trying to compost packaging, or just throwing it in the bin with all their other waste, or even worse, in the recycling?

The second part of the study was a mass participation experiment to test how well biodegradable plastics break down in compost heaps of all kinds. How long would it take? Would the plastics break down at all in normal, home composting conditions? Do some compost systems break down biodegradable plastics better than others?

Having these two well defined stages was vital to the success of the project. When your research team is thousands of people lending their time in all corners of the UK, changing any parameters mid-study could be a little tricky.

But knowing what data you want to collect is only the starting point.

The first part of the study (the survey) was relatively straightforward for people to do, but the second part asked a lot of its participants. People needed to put plastic labelled as biodegradable or compostable into their otherwise normal composting systems. They had to record the results periodically by taking photos, writing short descriptions and noting the scale of degradation. We had to make the process of sharing the results as easy as possible to maximise long-term participation and allow meaningful data to emerge.

The issue is that the easier an experiment is to complete, and the platform to use, the wider the likely participation and volume of data created. But the simpler you make the process for users, the less detail you can actually collect in your data.

In short, when it comes to citizen science, participation will fall as the complexity of your experiment and user interface increases. The trick is in finding the balance between usability and the type of data you need.

Creating a great user experience is key to bringing people on board with your study.

We knew, for example, that our system had to have a really simple way of interpreting the degradation of the material in their compost heaps. It also had to capture a level of detail that would be useful to Mark and Danielle. We implemented a 0–4 degradation scale. Each point on the scale is accompanied with an example photo and short description of what’s happening. This helps participants understand and interpret what they see in their compost. Users upload photos to accompany each rating, which allows the Plastic Waste Hub to do a quick visual check against the degradation rating and ensure consistency in the data.

If this crucial interaction was clunky, slow or poorly designed, then all the Plastic Waste Hub’s marketing efforts to engage the public would be wasted.

Speaking of...

The more people engaged in your study, the more data you can collect, and the more meaningful your results will be. Mark Miodownik has a relatively high profile in the UK, thanks to his work in broadcasting and other outreach activities. We’re sure that helped The Big Compost Experiment get on BBC’s Inside Science — once on launch and once a couple of months later to update on progress.

Mark was lucky to have a national platform to promote the experiment, but that alone isn’t enough. You also need to be sure that you’ve done all of your search engine optimisation (SEO) upfront, so that people who hear about you on the radio or TV can easily find you when they’re thinking back to it later. It helps to give your study a catchy name — but nothing too generic. “Big Compost Experiment” works because it’s specific, easy to remember and isn’t competing too much with the billions of websites already out there. It was fairly simple for us to prep the experiment’s website for people Googling “big compost experiment”, and even just “compost experiment”. People listening to the radio can just type in the name of the study and find our website fast. But if the study was called something more vague like “Plastic Experiment” it would have been much harder to stand out in search results.

Social is another obvious marketing tool for an experiment like this. Getting people with influence to link to your study can boost participation a great deal. Forums for niche communities can be a great help, too. There’s a forum for almost everything, and the people on them tend to be super-enthusiastic about their given subject. It’s worth taking the time to get to know these people, maybe even before designing your experiment.

As with any marketing, it helps to have a network of ambassadors to deliver your message far and wide. Knowing ahead of time who these people might be, could be a real benefit.

To encourage participation we built an interactive map to show the locations of participants. Linked to the map is a pre-moderated image gallery for people to show their composting experiments. The idea is that if people see lots of other people doing the experiment, they’ll be more likely to do it themselves. The photographs also contribute to the data, showing the type of compost system (who knew there were so many?), start date and experiment stage. The more people share, the better the study becomes. It also contributes to the spirit of openness in citizen science. As Mark said on Inside Science, “This is a citizen science experiment, so the results are available to all. You can look up and see who put what in, when, what their composter was, and what happened. Everyone can look at this data”.

We automatically send emails to participants, based on their own usual compost cycle, to keep them engaged and the data coming in.

Danielle produced the great illustrations that accompany the site. They help add to the feeling that the study is a community event, and give the website character and style.

The study has been very popular and — in a nice surprise — we had underestimated how much some people were willing to do. We acted on their feedback. The system was designed to carry out one experiment per participant, but now it can now handle multiple experiments. This seemingly simple change allows enthusiastic individuals to test different compost heaps, different plastics, different seasons and so on which will in turn create better data for more meaningful analysis.

It’s really exciting to see data coming in and to start analysing it. The sooner you can share any findings — even if they’re pretty vague — the better. The Big Compost Experiment’s six month results have been released. Sharing initial analysis keeps people engaged, especially in longer-term studies where you’re asking people to check in periodically.

Mark Miodownik and Danielle Purkiss went on BBC Inside Science again to do precisely this, and to encourage more people to get involved with the experiment.

We can’t wait for more updates on the Big Compost experiment. It’s a great project to work on, and if you’re interested in doing citizen science on the web, we’d love to help.