Statistics New Zealand has taken a collaborative and user-focused approach to developing the product and service mix for the 2006 Census.
Using the audience analysis methodology outlined in Content Management Bible (Bob Boiko), Statistics New Zealand set out to identify its audiences, with the intention of being better able to meet each audience's needs for using and understanding official statistics. The use of an audience model ensures that the appropriate products and services are produced to meet the particular needs of each audience.
Analysis of Statistics New Zealand’s audiences identified three broad audiences, shown in the inverted pyramid below.
Statistics New Zealand Audience Model
The public audience is made up of all members of society who want information in a personal capacity or on behalf of another organisation.
The professional audience is made up of professionals in the public, private and educational sectors who use and analyse statistics for the purposes of policy-making and other high-level endeavours.
The technical audience is made up of professional and academic statisticians working in the area of statistical methodology, plus expert data users.
top
In October 2004, the Census business unit of Statistics New Zealand carried out extensive surveying of and consultation with users to ascertain their needs for census data.
The review aimed to understand the needs of users in respect of the suitability of the products and services produced from the 2001 Census. Users were asked to provide feedback and suggest improvements to the content and functionality of the 2001 products and services. The review also aimed to identify and understand the major decision-making points of key internal and external users who need census data. The research recognised that users’ time requirements for census data would be critical to deciding when, and in what order, products are delivered.
Consultations were carried out with central and local government, researchers and academics, as well as Statistics New Zealand’s internal users. In addition, web-based questionnaires were distributed to gather feedback from a wider group of external users.
From this research, the outputs users require from the Census were identified, as well as the appropriate release dates for these outputs.
top
The Product Development and Publishing business unit of Statistics New Zealand subsequently carried out research to determine how these outputs would be delivered to users.
The research process adopted is based on methodology pioneered by Alan Cooper (author of About Face 2.0: The Essentials of Interaction Design), which is now regarded as accepted practice for understanding user needs. At the core of this methodology is the creation and use of personas.
Personas are "a precise descriptive model of the user, what he/she wishes to accomplish, and why" (Cooper). These models are represented as individual people; they are deliberately realistic and personalised with names, jobs and interests.
Ethnographic interviewing
As recommended by Cooper as the most effective way to gather good user information, the research involved conducting ethnographic interviews. These are directed discussions that involve observing users and talking to them about their needs when they use a product – for Statistics New Zealand, this was the Internet and print publications. The primary aim of the interview is to find out the users’ goals in their work, as well as to gain an understanding of their motivations and frustrations.
It was important that the interviews were conducted as discussions rather than as formal interviews because, as interviewers, we didn’t know enough about their domain to assess what was important to them. Through the interviewees showing and telling us about the current products and work methods they used, we were able to gain a better understanding of their likes and frustrations than if a set of fixed questions had been asked. Nevertheless, a guideline of topics to cover was used.
Also critical to success was to conduct the interviews where the interviewees carried out their work. Not only did this give us the opportunity to witness how they worked, it also gave us access to the environment in which they work. This provided us with a greater insight into the types of constraints they faced (for example, the age and capacity of their computer software) and, in turn, helped us to better identify their needs and goals.
Persona hypothesis
Prior to conducting interviews, it was essential to ensure that the interviewees encompassed the wide range of roles and organisational types within Statistics New Zealand’s three main audiences – public, professional and technical.
The persona hypothesis is a first attempt at defining the different kinds of users of a product within each audience. It attempts to address three main questions:
- What different sorts of people might use this product?
- How might their needs and behaviours vary?
- What ranges of behaviour and types of environment need to be explored?
For the public audience, Statistics New Zealand collaborated strongly with the producers of the Encyclopedia of New Zealand, and drew heavily on their research into the public’s access to the Internet. Subsequent research into the public audience focused on gathering more specific information from community groups, librarians and individuals.
For the professional and technical audiences, 30 interviews were undertaken with policy planners and analysts, journalists, managers, statisticians and researchers from central and local government, research and private sector organisations.
Following the interviews, write-ups of responses and observations were completed.
Mapping behaviours
Having identified the behavioural variables during the persona hypothesis stage, each interviewee was then assigned a number. Assigning a number to each interviewee not only ensured responses were confidential, it also removed the possibility of a mapping bias occurring between interviewees with the same job titles.
Each interviewee was then mapped against each behavioural variable. The precision of this mapping was not of critical importance; rather, it was important to focus on the placement of interviewees in relationship to each other.
Creating personas
Once all interviewees had been mapped, clusters of numbers began to emerge across multiple variables, indicating there was a meaningful behavioural pattern. Each cluster formed the basis of a persona, which were then developed further using the details from the interviewee write-ups.
Verifying the personas
At this point, the mapping of the behavioural variables was again checked to ensure there were no important gaps, or potential personas that had not been identified. The cast of personas was then checked for completeness and distinctiveness.
Statistics New Zealand has 11 personas (five public audience, four professional audience and two technical audience), which formed the basis for the 2006 Census products and services.
top
A series of internal workshops combined the research undertaken by Census and the personas created by Product Development and Publishing to determine the products and services mix.
The workshops drew on the expertise of subject matter areas for input. It was recognised that subject matter areas work heavily with external users of census data, and by drawing on their expertise it gave the workshops a wider and richer perspective.
The workshops centred on the personas’ needs and goals, which were used to articulate the various components of products and services that would best meet their needs - for example, the look and feel, delivery medium, and type and format of content.
Moving forward to the product design stage, being able to provide product developers with realistic 'pictures' of representative users will give them a powerful tool for realising user needs in their design work.
top
Extensive research into the time constraints and content needs of users, coupled with the development of personas, has enabled Statistics New Zealand to develop a range of products and services for the 2006 Census that will better meet user needs.
Cooper, A and Reimann, R (2003). About Face 2.0: The Essentials of Interaction Design. Wiley Publishing Inc, Indianapolis.
Boiko, B (2002). Content Management Bible. Hungry Minds Inc, New York, 498-503.
Lauren Wood
Product Development and Publishing
Statistics New Zealand
PO Box 2922
Wellington
New Zealand
Email: lauren.wood@stats.govt.nz
Printable version (235 KB) The downloadable file is in MS Word 2002 format.
Powerpoint presentation (525 KB)