Valuable insights from Homeless Link’s DataDive

Thursday, 17 August 2017 - 5:50pm

From 9am on Saturday 29 July until 5pm on Sunday 30th, with barely time to sleep in between, an amazing 80 data scientists and visualisation experts volunteered their time to help us explore some of our biggest data sets.

Group of data analysts

Before anything else, we need to say a huge thank you to everyone who took part - analysing our data, challenging us, and giving us new insights and fresh ideas. We really appreciate your generosity, and both StreetLink and In-Form DataLab will be better for your contributions.

What did we ask the experts to tell us?

1. Using StreetLink data from the last five years, we asked volunteers to explore who makes StreetLink referrals, what predicts how busy the service will be, and whether there are identifiable predictors of positive outcomes.

2. Using data from the In-Form DataLab, we asked volunteers to develop new insights into who uses services, the routes people take through services over time, and the outcomes they achieve.

StreetLink: What did we learn?

What predicts StreetLink referrals?

We learned that weather is a strong predictor of referrals to StreetLink - the search term ‘StreetLink’ was highly correlated with search terms related to household boilers in Google, indicating that people start to think about sending alerts for rough sleepers when it’s time to put the heating on again!

There is also a link with social media activity. Twitter is a driver of referrals, but content matters; certain words and phrases are more likely to prompt referrals than others. StreetLink is already thinking about how it can use this analysis to drive referrals and to plan the staffing of the team that manages the referrals.

What data is a predictor of outcomes?

Analysis indicated that environmental factors, such as the type of weather, were a predictor of positive outcomes, with referrals made in sub-zero temperatures most likely to have a positive outcome.

Importantly, the kind of information people included in their referrals was also found to predict positive outcomes. Referrals with a higher chance of a positive outcome include those that describe the location of the rough sleeper and their appearance, and who make their referral through the app. We will now go on to use this insight to inform the development of the new StreetLink app and website.

Volume of StreetLink referrals across London by Local Authority

DataLab: What did we learn?

We took away some fantastic tools that individual organisations can use to gain insight from their data, including an interactive tool that is able to explore data from the Homelessness Outcome Star.

Looking at the dataset as a whole, we learned that there appears to be a gendered difference in the services that people access, with fewer women accessing outreach services or direct-access hostels. A gendered difference was also noticeable in the routes that individuals take through services, with more women following ‘positive routes’ than men, who were more likely to move from a floating support service into a supported accommodation service. 

Warning!

We need to be careful about how we interpret these results as they are based on a relatively small dataset, and we must take more time to understand the data before we can come to any definite conclusions. The DataDive also showed us that we have more work to do before the In-Form DataLab can be used as a new source of information about the sector!

We also came away with some amazing visualisations of data from other homelessness datasets, like the gif you can see above.

Where do we go from here?

The StreetLink team got straight to work. The service is currently undergoing a redesign and learning from the weekend has been fed directly into this. You can expect to see the new and improved StreetLink this autumn.

Keep your eyes peeled for the next steps of the In-Form DataLab. We’ve got some great ideas about where we take the project next and will be reaching out to In-Form users in the coming months.

In the meantime, if you want to get in touch with us about the In-Form DataLab, feel free to email Maarten.Rooney@homelesslink.org.uk.

Talk To Us

Debra Hertzberg

Research manager

Debra is research manager at Homeless Link, coordinating research into homelessness and across other areas of social exclusion. 

Telephone: 020 7840 4422
Email: debra.hertzberg@homelesslink.org.uk