UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

40
Electoral Commission project TM

description

The Electoral Commission was concerned about whether the design of ballot papers was making it difficult for electors to vote accurately.Spurred by this, the Electoral Commission commissioned User Vision and Effortmark to conduct usability tests with a range of voters in Scotland, England, Wales and Northern Ireland.In their talk, Clare and Caroline will provide insight on how they approached the project and what they learned about ballots and about running paper testing.

Transcript of UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Page 1: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Electoral Commission project

TM

Page 2: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Clare Barnett and Caroline Jarrett

Clare is a usability consultant with User Vision

Caroline is a usability consultant / forms expert

Other people involved: Louise Ferguson advised the Electoral Commission on

usability issues Will Reburn of the Electoral Commission was our main

contact for this project Chris Rourke & Rob Van Tol from User Vision worked with

us.

Page 3: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

And now, a word from Caroline’s sponsor

Redish (2007)“Letting go of the words:writing web content that works”Morgan Kaufmann

Jarrett and Gaffney (2009)“Forms that work: Designing web forms for usability”

Morgan Kaufmann

Page 4: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

How would you cast your vote?

Page 5: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Outline: we tested ballots in four countries

Background to the testing How we approached the testing A brief description of the main findings

What makes voting hard or easy The detailed design findings

How we analysed the results Offline and online testing, differences and

similarities

Page 6: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

This project was a response to problems in 2007 After the 2007 Scottish elections, and other

elections, concerns grew about the usability of election materials.

The Electoral Commission commissioned usability testing of: existing voting materials, the use of registered party descriptions on ballot papers.

Focus groups were also organised by Brahm. The end goal was to feed into the development of

draft design standards creating consistency.

Page 7: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

The participants represented a cross-section of UK voters Location

Scotland, England, Wales, Northern Ireland rural, suburban, and urban locations

Age from 18 to 67

Voting experience never voted, occasional voters, frequent voters

Different voting methods Polling station & postal

Social classes Educational levels

from no qualifications to degrees

Nationality & first language Including Welsh speakers.

Page 8: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Outline: we tested ballots in four countries

Background to the testing How we approached the testing A brief description of the main findings

What makes voting hard or easy A selection of the detailed design findings

How we analysed the results Offline and online testing, differences and

similarities

Page 9: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

The test was an ordinary task-based test

Each participant had to complete a selection of voting tasks, varied by location. Two examples: In London: General election, Mayoral election, London

Assembly, European election, Local election by postal vote In Scotland: General election, Scottish Parliament,

European election, Local election

For each voting task: we asked who they were going to vote for we gave them the appropriate materials they voted we briefly discussed the experience of that particular vote we offered a short distraction task between votes

We had a ‘beauty competition’ of the materials to look at the finer details.

Page 10: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

We co-ordinated our approach

We created a generic crib sheet that worked for all the different voting patterns

We used the first test in London as a pilot Caroline facilitated Rob and the clients watched We debriefed immediately

We passed our experiences along: Rob tested in Wales Caroline tested in Cornwall Rob and Clare worked together in Scotland Clare tested in Northern Ireland

We all met in Scotland to debrief later.

Page 11: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

We used a distraction task between votes

As the voting process is a fairly repetitive task, we provided distraction tasks between voting.

Page 12: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Each session required skills in paper juggling

For a typical participant we had: 4 ballot papers 3 station posters 3 booth posters 1 set of postal instructions

Postal voting statement (or declaration of identity in N Ireland)

Textual instructions Illustrated instructions 2 envelopes

our generic crib sheet template for notes.

Page 13: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Outline

Background to the testing How we approached it A brief description of the main findings

What makes voting hard or easy A selection of the detailed design findings

How we analysed the results Offline and online testing, differences and

similarities

Page 14: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Two big factors: voting experience and the voting task(s) Participants found many plausible (though

incorrect) interpretations of the materials. We found that:

Voting experience is important. The structure of the election has a major effect. The complexity of the postal voting process

caused difficulty.

Page 15: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Voting experience is crucial to voting success If offered the choice,

even new voters skip the instructions

If asked to read the instructions, new voters read more carefully

Even if they had read the instructions, new voters became confused and some made mistakes Extract of Ballot Paper completed by a new voter

Page 16: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Participant deals with a complex election as if it were an ordinary one Example

Page 17: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Voting is harder than you realise:He votes the way he has done for years

Page 18: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Am I voting for a candidate or a party?

This participant tried to vote for an individual

The instruction to ‘vote for a party or an individual’ can be misinterpreted to mean ‘vote for an individual within the party’.

Page 19: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Why do I have more than one vote?

London Mayor Voter can indicate first and second choice “Can I vote twice for Boris?” “Do I have to vote for someone else?”

Scottish Local Election You can vote for as many as you like “Why would I vote for more than one?”

Welsh (& English) Local Election “Why do I have so many votes?” What happens if the voter does not use them all?

Page 20: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Outline

Background to the testing How we approached it A brief description of the main findings

What makes voting hard or easy A selection of the detailed design findings

How we analysed the results Offline and online testing, differences and

similarities

Page 21: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Why have the candidate’s addresses on the ballot?

Some participants thought this was unnecessary Others worried about the potential danger to the

candidate Some were concerned that it might sway floating

voters

Page 22: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Logos are important. Some of them are missing Some participants didn’t vote for parties without

logos as they didn’t look “official”.

Page 23: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Numbers next to candidates are not useful

Numbering on the ballot is unhelpful There is a risk of circling the number instead of voting

properly.

Numbering within the party is also unhelpful Potential for error in voting for a candidate rather than a

party.

Page 24: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Candidates names are in an unfamiliar format Some ballots have:

SURNAMEFirst names

Other ballots have: SURNAME

Full name

And other ballots have even more:

Page 25: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Party names are important, but sometimes hard to find Scottish Conservative & Unionist Party Conservatives Conservative Party Welsh Conservatives Welsh Conservative Party The Conservative Party

Scottish Labour Party The Scottish Labour Party Labour Party

Party names are included in the brackets with the candidates name and address.

Page 26: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Instructions on the ballot were often poor

Some ballots did not have important instructions: Title of the ballot How to vote

Some ballots had instructions, but not easy to see London elections white-on-black instruction

If participants saw this instruction then they liked it:

Page 27: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

English and Welsh should be kept aligned

Welsh and English together double the information

There was no consistency over which came first Participants do not mind which, slight preference for

Welsh first

They do not want to see separate English and Welsh forms

Translations into Welsh were not exactly the same Mixing Welsh and English is confusing

Page 28: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Example of mixing Welsh and English

Page 29: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Speak in voters’ language

Give it a simple title. Don’t bunch important

instructions into one sentence.

Don’t double information by repeating it.

Page 30: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Polling station notices have instructions that are hard to understand What does ‘spoil a ballot’ mean? What’s a compartment? What’s this about showing your vote to the

presiding officer? Instructions have numbers alongside them (good)

but often more than one item included for a number (bad)

Why tell voters to leave the polling station immediately?

Page 31: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

The poster in the booth should tell you what to do in the booth If they need it before, put it on a poster If they need it in the booth, tell them in the

booth What type of mark to make

“Do I use a cross or a tick?”

The rules about the numbers of votes you have What to do if you make a mistake.

If they need it after voting, put it after Fold/don’t fold, Show to member of staff.

If they don’t need it at all, scrap it! “Leave the voting station immediately”

Page 32: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Outline

Background to the testing How we approached it A brief description of the main findings

What makes voting hard or easy A selection of the detailed design findings

How we analysed the results Offline and online testing, differences and

similarities

Page 33: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Collating analysis was a joint task

There were 3 testers, 45 participants and 4 countries. We entered our findings per country into an Excel

spread sheet Caroline came to Edinburgh

We each had 30 mins to talk about our particular test results for each country

This allowed us to see whether issues were country specific or UK wide

Each issue was written onto a post it and placed on a white board under country specific or UK wide

We then removed any duplicated issues from the board

The Electoral commission wanted a summary presentation of our main findings, followed up with a report detailing a full list of the issues and recommendations.

Page 34: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

We wrote the full report with client involvement The Electoral Commission wanted to make the report a

public document. Due to experiences with previous reports, the Electoral

Commission did not want any figures referred to in the report. “In a small qualitative study, it is important to focus on the

problems that were identified by participants rather than the numbers who found those problems.“

“We also did not consider how difficult or easy our recommendations may be to act on for legislative or any other reason.“

We split issues into categories. These were agreed with the Electoral Commission.

The post-its from the analysis meeting were used, and split into categories, flowing as much as possible with the journey that voters take.

Page 35: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Full report

Electoral commission website: http://www.electoralcommission.org.uk/doc

ument-summary?assetid=77687

Page 36: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Outline

Background to the testing How we approached it A brief description of the main findings

What makes voting hard or easy The detailed design findings

How we analysed the results Offline and online testing, differences and

similarities

Page 37: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Offline and online testing: similarities

The process followed was the same for online testing.

Flapping about of paper is similar to users being unable to find information online: Users are having trouble finding something.

People skipped instructions, even if they were only a few words long.

Users seemed to ‘get into’ the idea of the test quickly and treat it as realistic.

Page 38: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Offline and online testing: differences

“Feel appeal” – you actually interact with the materials You can draw straight onto the materials. Paper is an physical product and can be bulky.

On paper, signatures really matter. Recording is harder:

One venue: good quality video – of the participant’s head

Other places: used a web cam, but quality not very good.

Page 39: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

What next?

The Electoral Commission has published their guidelines – Making your Mark:

http://www.dopolitics.org.uk/making-your-mark

Page 40: UKUPA Jan 10 Clare Barnett and Caroline Jarrett: Election Ballot Usability

Question time

Caroline [email protected]

01525 370379

Clare [email protected]

0131 225 0850