2 из 40

These slides are set up so that they can be printed back to back (two/four sided) to give:

A short hand overview about when to use each method
A summary of the method, what it’s good for and limitations (linking to other slides in this pack)

3 из 40

Choosing research methods

When you need to think about which method is best in theory and in practice

4 из 40

Choosing Research Methods

Providing a rationale for the methods you choose to use and how you employ them.

What are your research goals? If you are looking to influence experts or policy makers, quantitative approaches will add weight to your findings. If you are looking to understand problems, inform innovation or develop a prototype, look at qualitative methods or user research
What are your research questions? If they begin with ‘explore’ or ‘what’ look at qualitative methods (talking). If they begin with ‘identify’ or ‘why’ look at quantitative (see guide to research questions )
What research traditions exist? You may choose to follow or challenge them. Think about whether you want your research to be noted for its quality and robustness or creative approach and unique insights
What are your/your teams skills? You may not be an expert in the most appropriate method so consider asking for other team members or commissioning out research
Who are you research participants? Think about your relationship to participants (especially if you are doing qualitative research) and how they will respond to you and the method. Consider if they are often consulted or surveyed and whether if could be helpful or unhelpful to stick with their comfort zone or not.

5 из 40

Using online tools

When you need to decide which tools to use for research

6 из 40

Using online tools

What to think about when choosing a tool to conduct research

What’s the cost to the research quality ? Most tools are ‘freemium’, use a basic version for free. BUT these are designed to annoy you to pay to do good research. Consider privacy settings, data access, storage and value for money. Survey tools will have no option to filter participants (if yes/no answer this q), a 10Q limit, no branding. Mapping/visualisations are published online and open source tools aren’t always user friendly
Start with user needs, understand the context and think about everyone. Consider what technology they have, how they will access the tool and what they need to do this. Do they have internet, data, time?
Be creative: Online tools may not be designed for research, but Google Forms, Trello, Workflowy and Slack are all valuable collaboration tools. Twitter and Facebook polls may increase participation in research. However, think about what they are missing, what they can’t do and pilot your analysis approach first
See what’s out there: This online sheet of Applied Social Research Guides and Resources includes a list of online tools for research and evaluation to test. Those widely used for your research method or sector are likely to be the best starting point. Some tools allow you to do research (see Tags for Twitter data capture), analyse it or present it in new ways (see Raw Graph s for data visualisation)

7 из 40

Contents: Methods summary

Structured Interviews : When you want to gain a broad range of perspectives about specific questions
Semi-Structured Interviews : When you want to gain in-depth insights about broad questions
Unstructured Interviews : When you want to gain in-depth insights about a complex research topics
Telephone Interviews : A tool for when you want to interview people quickly and easily
Guerilla Interviews : When you want to carry out user research or explore general perspectives quickly
Contextual Interviews : When you want to understand actions and particular experiences indepth and in context
Focus Groups : When you want to understand shared experiences and different perspectives
Participant Observation : When you want to ‘learn by doing’ or observe social interactions and behaviour
Ethnography : When you want to experience social practices, interactions and behaviour with minimal influence
Surveys: When you want to generate numerical data about the scale of people’s opinions and feelings
Mixed Methods: When one method cannot fully answer your main research question
User Research : When you want to learn about the behaviours and motivations of your target audience
Service Design Research : When you want to design a service to meet people’s needs.
Content Analysis : When you want to understand public discourse through secondary or online data
Workshops : When you want to engage stakeholders in research, generate ideas or codesign solutions
Usability tests : When you want to test prototypes or learn about problems with an existing service

8 из 40

Applied social research: A curated online sheet of Applied Social Research Guides and Resources
Surveys : Guide to creating questions here and here , build on existing data/questions , analysis guide
Interviews : A nice overview here which includes how to structure an interview
User research : The GDS for intro guides and DisAmbiguity blog
Service design: This is Service Design Doing has great tools and formats for workshops

Inspiration for emerging research methods and creative formats for research

Ethnography and mixed methods presented well: Ikea At Home Report
User mapping techniques as a social research method NPC Report
User Research to understand domestic abuse experiences and the potential for technology Tech Vs Abuse
Using Twitter data for social research Demos
Data visualisation as a tool for research communication - Nesta data visualisation and Women’s Aid Map
Data journalism and data storytelling - Guardian reading the riots
An online games to shift perspective on a social problem - Financial Times Uber Story
Content analysis to map trends - Nesta analysed creative skills in job adverts
Issue mapping online - networks of websites and people on Twitter - Warwick University Issue Mapping

9 из 40

When you want to gain a broad range of perspectives about specific questions

10 из 40

A conversation with a set structure (a script of fixed questions) and specific purpose. Can be a method to undertake a survey or called a ‘directed’ interview.

Asking standardised questions across many participants makes data easier to analyse and compare
Giving participants a clear guide about what you want to learn from them
Topics that would be too complex to capture in a questionnaire tick box/short response
Respondents with limited time, who want to consider responses in advance or do not want to write
The quality of the interview is less dependent on the interviewer and their rapport with the interviewee

Limitations (and how to avoid or what to consider instead)

The structure prevents participants from bringing in other ideas (consider semi-structured interviews )
Whilst quicker to conduct and analyse than semi-structured interviews, they are still resource intensive and only possible to do with limited numbers of people (consider questionnaires online - see surveys )

11 из 40

When you want to gain in-depth insights about broad questions

12 из 40

Conversation with a structure (set of open questions) and clear purpose. Also called directed interviews.

Exploring a range of perspectives on research questions, engaging experts and getting buy-in to research
Gaining in-depth insights about how people feel or interpret complex issues
Topics which are sensitive, difficult to express in writing or to articulate views about in a survey
Allowing participants to respond in their words, framing what they see as important

Quality can depend on interviewer skills and put people on the spot (consider setting topics in advance)
The set-up affects the quality of engagement and discussion (consider location, relationship with the interviewee and whether you should do a face to face or Telephone/Online interview )
Time consuming to do, analyse and compare (consider Structured Interviews or Focus groups )
Can lack validity as evidence (consider Surveys )
Explore what people say, think and remember, not what they actually do (consider Participant Observationcontextual interviews or User Research ) or shared perspectives (consider Focus groups )
Easy to provide too much structure and prevent open exploration of a topic (see unstructured interviews )

13 из 40

When you want to gain in-depth insights about a complex research topics

14 из 40

A loosely structured open conversation guided by research topics (also called non-directed interviews)

Very exploratory research and broad research questions
Letting the participant guide the interview according to their priorities and views
In-depth and broad discussion about a person's expertise, experiences and opinions

Participant can feel like the they are not saying the ‘right’ thing (explain technique and rationale well)
Whilst useful for expert interviews, an unstructured approach can give the impression that the interviewer is unprepared, lacks knowledge or the research purpose is unclear (consider semi-structured interviews )
Interviews are longer, resource intensive and only smaller numbers are possible (consider focus groups )
Generates in-depth insights that are difficult to analyse and compare
A lack of structure can encourage participants to focus in-depth on one thing they are positive about or know very well in-depth (consider using desk research to inform the interview topics)

15 из 40

When you want to carry out user research or explore general perspectives quickly and easily

16 из 40

An ‘impromptu’ approach to interviewing, often talking to real people on the street or at a key site

Gaining immediate responses to a tool or design and insights into a problem
Informal method means participants can be more relaxed and open
Speaking to a lot of people, simply, quickly and cheaply about one key question
User research and user experience of interacting with digital products

Limitations (and how to avoid or what to consider instead)

Speaking to people for convenience (users are available in a single place and time) introduces sample bias (but you can add more targeting and profiling of participants, see the Guide to Sampling )
The lack of formal structure can mean that you miss important questions or insights
Findings are often unreliable and not generalisable because they rely on a single type of user
Difficult to understand complexity or gain contextual insights

17 из 40

Telephone / online interviews

A tool for when you want to interview people quickly and easily

18 из 40

Telephone or Online interviews

A tool to conduct an interview (it is not a method in itself) which is not in person/ face to face

Conducting interviews without the costs of travel and meeting time (often shorter)
Expert and stakeholder interviews, when you already know the participant well or they are short of time
Taking notes and looking up information whilst interviewing is less disruptive than in person, easy to record
Sending informed consent information and interview questions in advance

Can be difficult to undertake an engaging interview (hard to build rapport on the phone)
Often need to be shorter and put alongside other meetings

What method are you using?

Structured interviews : When you want to gain a broad range of perspectives about specific questions
Semi-structured interviews : When you want to gain in-depth insights about broad questions
Unstructured interviews : When you want to gain in-depth insights about a complex research topics

Further guides to Interviews : A nice overview here , including how to structure an interview

19 из 40

When you want to understand actions and particular experiences in-depth and in context

20 из 40

Interviews conducted with people in a situational context relevant to the research question; also known as contextual inquiry.

Understanding what happens, experiences and emotions whilst interacting with a tool, service or event.
Easier for research participants to show rather than explain, participants are active and engaged
Uncover what happens, what people do, how they behave in the moment, rather than how they remember this and give meaning to these responses later.
Open and flexible method giving depth of insights about a tool or specific interaction

Limitations (and how to avoid or what to consider instead)

Time and resource intensive for the researcher
Each context is unique - making it difficult to generalise from or to answer broader research questions about experiences (consider semi-structured interviews )
The researcher influences the interactions and events (consider ethnography or participant observation )

21 из 40

When you want to understand shared experiences and different perspectives

22 из 40

An organised discussion with a group of participants, led by a facilitator around a few key topics

Gaining several perspectives about the same topic quickly
Research contexts and topics where familiarity between participants can generate discussion about similar experiences (or different ones) which may not arise in a one to one interview
When attitudes, feelings and beliefs are more likely to be revealed in social gathering and interactions
Including tasks and creative methods to elicit views (e.g. shared ranking of importance of statements)

Difficult to identify the individual view from the group view (consider semi-structured interviews )
Group dynamics will affect the conversation focus and participation levels of different members
The role of the moderator is very significant. Good levels of group leadership and interpersonal skill are required to moderate a group successfully.
The group set-up is an ‘artificial’ social setting and discussion (consider Participant Observation )

23 из 40

When you want to ‘learn by doing’ and observe social interactions and behaviour

24 из 40

Participant observation/ shadowing

The researcher immerses themselves in lives of participants as an ‘observer’ of their behaviours, practices and interactions. A type of ethnography. The people being observed know about the research.

Understanding everyday behaviours, interactions and practice in the context that they occur
Gaining an intuitive understanding of what happens in practice and what this means for those involved
Allowing research participants to show you what they do, when they can’t describe and remember this well
Establishing topics for further investigation through more structured or focused research methods

If explicit (shadowing for example) the research situation is still ‘artificial’
Your audience may not respect it and can be difficult to generalise from (consider mixed methods)
The quality of the data is dependent on the researchers’ skills and relationships with participants

25 из 40

When you want to experience social practices, interactions and behaviour with minimal influence on what happens

26 из 40

The systematic study of a group of people or cultures to understand behaviours and interactions. The researcher becomes an ‘insider’. It is a way of presenting research findings, as well as a method, which can include participant observation, document analysis and visual methods.

When you need to be an ‘insider’ to fully access the research context (such as organisational cultures)
Presenting how everyday behaviours, interactions and practice occur in context
Gaining an in-depth knowledge of your research context, participants and social relationships
When little is known about a research context or topic

If covert (at a conference or workplace for example) it has implications for informed consent
If explicit (shadowing for example) the researcher’s presence can affect the interactions and findings
The quality of the data is dependent on the researchers’ skills and relationships with participants

Example use case : Ikea At Home research study to understand how people feel about their home

27 из 40

When you want to generate numerical data about the scale of people’s opinions and feelings

28 из 40

A process of systematically collecting information from a large number of different people. Responses are summarised as statistics (online surveys automate this analysis for you).

Targeting specific types of research participant and providing data about their views
If designed well, they can be quick, simple and non intrusive for research participants
Findings can have more credibility than other methods because of their breadth
Describing, measuring and understanding (a basic questionnaire)
Statistical analysis, modelling cause and effect (large scale survey designed to represent the population)

Can raise more questions about what happens and why, lack depth of insight (consider mixed methods )
Hard to design well and require a lot of time upfront and data skills to analyse the results
Low completion rates and people feel ‘over surveyed’ (consider incentives )
Assumes people will be honest and sufficiently aware of the research context to provide credible answers.

Further information: A great guide to creating questions here and here , build on existing data/questions here

29 из 40

When one research method cannot fully answer your main research question

30 из 40

Combining different methods to answer your research questions, can be a mix of quantitative or qualitative methods or both. It may mean working with different types of data, research designs or being part of a research team (covering different research disciplines)

Overcoming the limitation of relying on a single research method or approach
Triangulating findings (i.e. using an additional method) can give them more validity
Accessing different types of research participants
A more holistic understanding about how, why and the extent to which something happens
Answering different types of research questions about frequency and perceptions
Giving findings more validity and influence because of the range of data and insights

Requires a broader range of skills and more time to deliver, analyse and report on
Research design must have strong sequencing (when each method is used and analysed , why) to make the most of a mixed methods approach - not always possible in a tight timescale or short research project

31 из 40

When you want to learn about people’s needs, behaviours and motivations for using a service

32 из 40

A research approach employed to understand users and their needs, motivations and behaviours, primarily to inform service design.

User-centered design processes which look to ensure services meet the needs of their audience
Gaining specific insights into how a person interacts with a digital tool or service
Exploring general needs, behaviours and motivations for a specific target group using a range of services

Focus on a tool or service can prevent wider analysis, relevance and applicability
Research can lack credibility due to small numbers, set up, documentation (often highly specific focus)
Can overlook those who do not use a service for a whole range of reasons

User research involves any method which looks at who users are, the problems they face, what they are trying to do and how they use existing services. This can create user personas, user journeys and user experience maps. It largely includes qualitative research methods.

33 из 40

When you want to design a service to meet people’s needs, including planning, organising, infrastructure, communication and components)

34 из 40

A research approach employed in the activity of planning and organising of people, infrastructure, communication and material components of a service, in order to improve quality and interaction.

User-centered design processes which look to ensure services meet the needs of their audience
Gaining a holistic picture of all components (infrastructure, people, organisations, culture) affecting how a person interacts with a service

Focus on a tool or service can prevent wider analysis, relevance and applicability
Research can lack credibility due to small numbers, set up, documentation (often highly specific focus)
Can overlook those who do not use a service for a whole range of reasons

Service design often begins with user research but participants in research include all those involved in delivering (not just using) a service, such as employees and stakeholders in an organisation as well as looking at the context and system which affect how a service works and its effectiveness

35 из 40

When you want to understand public discourse through secondary or online data

36 из 40

A systematic process of classifying and interpreting documents, text or images to analyse key discourses (their meaning) or to quantify patterns (such as word frequencies). This can be done manually or it can be automated.

Exploring the focus of messages, text or imagery and change over time
Secondary data sources, such as archives, online social media data (such as Tweets) and news articles
Gaining a qualitative or quantitative insights about key messages

Focuses on public and documented interpretations of events and experiences
Documents are not exhaustive and not all are accessible (or available online/freely)
Qualitative coding is time intensive to manually classify, reliant on researcher interpretation
Automated coding for key words can miss nuances and difficult to produce meaningful findings

Documents and other sources�What it is:�Any written, audio or visual materials that already exist. This could include documents relating to the charity’s activities, like case management data or previous research that has been conducted. It also includes digital materials. Analysis of these documents is best done in conjunction with interview or collecting group primary research data. It is one way of ‘triangulating’ different data sources to increase understanding.�Good for:� Avoiding unnecessary primary data collection. The qualitative information you need might be found in existing documents. For example you may have some qualitative data from key workers’ case notes or management systems. Similarly, media articles about a particular topic can be useful, or you may want to analyse local strategy documents to show variation in attitudes or services.� Supplementing primary (conversation) data, to help explain results.� Providing background information to help inform primary research questions.�Limitations:� It may be as time-consuming as primary research because although this data is already available, collecting and analysing it systematically is still important.

37 из 40

When you want to engage stakeholders in research, generate ideas or codesign solutions

38 из 40

A tool to undertake research. It is an interactive session, often taking a full day, in which research participants sor stakeholders work intensively on an issue or question. The process can combine elements of qualitative research, brainstorming or problem solving.

Engaging stakeholders - building empathy with and understanding of research findings
Understanding problems or prototyping solutions, linked to user research and service design approaches
Participatory research, allowing participants to shape agendas and outcomes
Creative, collaborative and engaging activities to build rapport and understanding with participants
Participatory design, enabling participants to co-design solutions which work for them

Highly dependent on the right people attending and the facilitation skills
Can be a lot of time and effort to coordinate a workshop effectively and analyse findings
The immersive and collaborative environment makes it difficult to document effectively
Collaborative solutions may duplicate existing problems or solutions

39 из 40

When you want to test prototypes or learn about problems with an existing service

40 из 40

A user research method where you watch participants try to complete specific tasks using your service. Moderated testing involve interaction with the research participant, asking them to explain what they are doing, thinking and feeling. Unmoderated testing is completed alone by the participant.

Identify any usability issues with a digital service - for example, problems with the language or layout
Seeing if users understand what they need to do in order to complete designated tasks
Generating ideas to improve a prototype of existing digital service
Assessing user experience

Limitations (and how to avoid or what to consider instead)

Focus is not on ‘natural’ use (consider contextual interviews , participant observation , ethnography )
Data is about a specific design and interaction with a tool at that moment
Findings cannot be generalised or applicable more broadly to understand users and behaviours