Christina Li

Feb 6, 2018

5 min read

How Google Map can help reduce your research bias

A recent research article by IDEO suggested mapping out each research location will allow you to see where your potential research bias is, i.e., you haven’t conducted research in just easy to reach locations. I’ve recently applied that approach for a public sector client in the UK.

For the last 10 months, we have been conducting research across 3 phases: Discovery (4 sprints), Alpha (4 sprints) and Private Beta (11 sprints and counting) around the UK.

The team is based in London and we were conscious that other parts of the UK may have different attitudes and behaviours when using government services.

Designing a service based on research findings in one geographic location wasn’t going to be helpful in the long run. So, we’ve made the effort to travel as much as we could to conduct research, but have we gone out of London enough? Have we reached different communities? I certainly feel that I have been on the trains a lot- each sprint (every fortnight) I am out of London 2 days doing research.

Now, it’s time for some hard data.

Research locations mapped across 3 key project stages (left to right): Discovery, Alpha and Private Beta

What have we learnt?

The map is visual and a great conversation starter

I mapped out the locations as an experiment to see if we had any bias in our research locations, and then I shared the map with the product owner.

For awhile I’ve been tracking the research locations on a page, so having the map in front of us shifted our conversation to look at where we have visited and where we haven’t. Instead of talking about number of participants we had spoken to, we were talking about where those participants were and if they were all concentrated in one area.

The map allowed us to review our future research plan (more on that in a bit).

Google Maps is a great tool to plot locations, but there are limitations

I used different ‘layers’ for the different phases of the research, so I could get a quick glance of all the locations in one go. The locations in one phase was highlighted with the same colour and icons shape (e.g., Discovery in pink with square vs Alpha in blue with circle), and each location represented around 5–6 users.

However, the ‘layers’ built on top of each other and covered up any previously mapped data. You can’t move the layers to a different order, so if you wanted to see your Discovery data on top of Alpha data you would have to remap everything! It’d also help to think about the order you wanted your layers to build up. In this instance, I went with a chronological order.

Another limitation is that you can’t map the same location twice across different layers — for instance, sometimes we visited the same participant over different phases to observe repeat usage and see if service maturity helped solve some of the earlier identified problems.

(This was my first time using Google Map this way, so do let me know if I could re-order the layers, or do something differently).

All research locations shown across multiple layers

We’re slightly London-centric in our research despite our efforts

Once this was all mapped out, I could see that the most number of places we conducted research in was London. So, despite our effort we still had a slight tendency to stay in the capital, or to visit another big city like Manchester.

One of our challenges is to consider how we can reach other cities or towns easily on public transport, and review our research recruitment strategy. How do we recruit people in ‘harder to reach’ locations who are potential users of our service? Do we have local contacts (e.g., a community centre) we can liaise with? What methods can we use to recruit successfully?

The most number of places we conducted research in is London.

The map highlighted areas we haven’t visited yet — we’re planning them now

Looking at the map it was very obvious we haven’t reached Scotland and Northern Ireland, and that was deliberate as the legislation we’re working with cover England and Wales only, and Scotland and Northern Ireland would have their own policy or legislation respectively.

Having said that, when I shared the map with the product owner, our conversation shifted from just recruiting for the next sprint of research to where could we do research in to fill our users knowledge gap. We also thought about if there were scenarios for people in Scotland or Northern Ireland to use the digital service even though the legislation might work differently.

We’re now talking about updating our upcoming sprints research plan to target specific locations so we can reach those ‘empty’ spots on the map.

We’ve reached more areas than we thought we could in 10 months

Yes, there are places we haven’t visited and perhaps places we have visited too many times. But, having seen all the locations on one map, we were quite pleased with our effort for the last 10 months.

We were pleasantly surprised that we managed to go to Wales a number of times- and thinking back to those trips, they were the ones that gave us a lot of insights and helped us refine our service too. We also went quite north to Newcastle (on a day trip) and we learnt a lot about people’s behaviours to the work we’re doing.

Now… more travels on the trains this week! 💪

How are you keeping track of your research locations? Anyone else mapping locations? Get in touch!

I founded Melon Experience Design to help clients gain better research insights from their users and make a social impact, get in touch if you would like to work with us.