Block 2 - Unit 2: Establishing an initial set of requirements Flashcards
What are we trying to achieve in the requirements activity? (2)
Identify needs:
Understand as much as possible about users, their work, and the context of that work, so the system under development can support them in achieving goals.
Establish requirements:
Produce, from needs identified, a set of stable requirements than form a sound basis to move forward into thinking about design.
Importance of getting the requirements right.
Cost of fixing errors late in the software development cycle is significantly higher than during the requirements activity.
Requirements gathering / capture?
Implies requirements exist out there and we just need to get hold of them.
Requirements elicitation?
Implies ‘others’ know the requirements and they just need to tell us.
Even if they ‘have’ them, they may not have articulated them yet, or have explored them in enough detail.
Requirements analysis?
Investigate and analyse an initial set of requirements that have been gathered, elicited or captured.
Important step, as interpretation of facts (rather than facts themselves) inspires the design.
Requirements engineering?
Recognises that developing a set of requirements is an iterative process of evolution and negotiation, and needs to be carefully managed and controlled.
‘Establishing’ requirements?
Choosen to represent the fact that requirements arise from data gathering, analysis and interpretation activities and have been established from a sound understanding of the users’ needs.
Also implies requirements can be justified by, and related back to, the data collected.
An aim of the requirements activity.
To make requirements as specific, unambiguous and as clear as possible.
Functional requirements?
Say what the system should do.
Fundamental to understand these for an interactive product.
Non-functional requirements?
Describe the various constraints there are on the product.
Can be technical (eg. needs to interface with another system);
or non-technical (eg. needs to support a particular type of user).
Examples of non-functional requirements. (4)
Must be able to run on a variety of platforms.
Target platform is expected to have at least 1 GB of RAM.
Must be delivered in 6 months time (constraint of development activity rather than product).
Interactive products in general - physical size, weight, colour and production feasibility.
Data requirements? (6)
Type.
Volatility.
Size / amount.
Persistence.
Accuracy.
Value.
Environment requirements?
Context of use - circumstances in which interactive product is expected to operate. (4 aspects)
4 aspects of environmental requirements.
- Physical environment.
Lighting, noise, dust, etc., protective clothing worn?, crowded. - Social environment.
Collaboration / coordination.
Eg. data shared? Synchronously or asynchronously?
Physical location of other team members. - Organisational environment.
Eg. how good is user support? Facilities / resources for training? How efficient / stable is communication structure? - Technical environment.
Eg. What technologies will the product run on or need to be compatible with?
What technological limitations might be relevant?
User characteristics?
Capture the key attributes of the intended user group - the properties of the users that impact on ID.
Some key user characterisitics.
Abilities and skills:
Novice - step-by-step instructions, prompting, constrained interaction.
Expert - flexible interaction with more wide-ranging powers of control.
Nationality, education, preferences, personal circumstances, physical or mental disabilities.
User profile.
The collection of attributes of a class of ‘typical user’.
One product may have several different user profiles.
Persona?
Rich description of typical users of the product under development that designers can focus on and design for.
Precise, credible details helps see personas as real potential users, and hence as people they can design for.
Details included in a persona?
Unique set of goals (inc UX).
Skills, attitudes, tasks and environment (detailed and specific).
Name, photo, personal details (leisure activities etc).
MoSCoW?
Prioritisation - ‘Must have’; ‘Should have’; ‘Could have’; and ‘Won’t have right now’.
4 issues for success of data gathering / recording sessions.
Setting goals.
Relationships between collector and provider (of data).
Triangulation.
Pilot studies.
Main reason for gathering data?
Glean information about something.
Eg. understand how technology fits into family life;
identify which icon best represents ‘send email’.
Why set goals for data gathering?
Many reasons for gathering data - need to identify specific goals.
Goals will influence the nature of sessions, gathering techniques and the analysis to be performed.
Once goals set, you can concentrate on what data to look for and what to do with it then.
Purpose of informed consent forms.
Gatherer - wants to know data can be used in analysis, presented to interested parties and be published in reports.
Provider - knows information will not be used for other purposes, or in a context that would be detrimental.
Children (parent signs) - no threatening / inappropriate / embarrassing questions.
Triangulation (def, example and reason)
Strategy of using more than one data gathering technique to tackle a goal, or using more than one data analysis approach on the same set of data.
Eg. observation to understand context of task performance, interviews to target specific user groups, questionnaires to reach a wider population, and focus groups to build a concensus view.
Provides different perspectives and corroboration of findings across techniques, therefore leading to more rigorous and defensible findings.
Pilot studies.
Small trial run of main study to make sure method is viable.
Data gathering participants are usually unpredictable, even with careful planning.
Potential problems can be identified and rectified.
Eg. equipement, instructions, questions.
Comments on pilot studies. (2)
Can use colleagues / peers for pilot study if difficult to find participants - quick and cheap.
Anyone involved in the pilot study can’t be involved in the main study - they will know more about it, which can distort the results.
Most common data recording methods. (4)
Taking notes.
Audio.
Photos.
Video.
Other data recording methods.
Questionnaires and diary notes ‘self-documenting’ (participant completes so no further recording needed).
Interaction logs usually generated automatically.
Factors affecting choice of data recording method.
Context, time available and sensitivity of situation - choice impacts on how intrusive data gathering is.
Most settings - audio, photos and notes sufficient.
Sometimes video is essential to record in detail intricacies of the activity and its context.
Notes plus still camera.
Notes - can be difficult / tiring to write and listen / observe; easy to lose concentration and bias to occur - can help to have someone else to do it.
Handwritten notes need to be transcribed - can be first step in data analysis.
Digital images can easily be collected (with permission).
Photos / sketches can be used to capture other images and docs needed.
Audio plus still camera.
Interviewer can pay more attention to interviewee, and is less intrusive than video.
May only need to transcribe sections - many studies don’t need high detail.
Audio can be supplemented wit photos of artifacts / events / environments.
Video.
Can be intrusive.
Attention is focused on what is seen through the lens - miss things out of view.
Sound can often be muffled - air con, wind, etc.
4 main types of interview.
“Conversation with purpose.”
Open-ended / unstructured.
Structured.
Semi-structured.
Group interviews.
(First 3 named according to how much control the interviewer imposes by following predetermined questions).
Choice of interview choice (and examples)
Depends on purpose, questions addressed and stage in lifecycle.
Eg. 1st impressions on new design - open-ended.
Feedback on particular design feature - structured or questionnaire. (More specific).
Unstructured interviews.
Exploratory and more like conversations, often going into considerable depth.
Open questions:
- no particular expectation on format / context of answers.
- used to explore range of opinions
Interviewee free to answer as fully or briefly as they want and both parties can steer the interview.
Some further points on unstructured interviews. (2)
Advisable to have a plan of main topics to be covered. (Having on agenda is different than being open to new information).
Balance between making sure answers to relevant questions obtained, and following new lines of enquiry not anticipated.
Benefits and cost of unstructured interviews.
Generates rich data - gives deep understanding of topic, and often interrelated and complex.
May mention issues not considered.
Cost:
A lot of unstructured data takes time to analyse.
Can’t replicate process as each interview takes on its own format.
ID and unstructured interviews.
Usually no attempt to analyse each interview in detail. Instead, use notes / audio to later go back to find main issues of interest.
Structured interviews.
Predetermined questions similar to a questionnaire.
Useful when goals are clearly understood and specific questions can be identified.
Questions typically closed - set of predetermined answers - and best if short and clearly worded.
Standardised - same wording and order with each participant.
Semi-structured interviews.
Combines features of (un)structured interviews, using open and closed questions.
Consistent - basic script for guidance, so same topics covered.
Starts with preplanned questions, then probes to say more until no new relevant information emerges.
Some points for conducting a semi-structured interview. (5)
Shouldn’t pre-empt an answer, eg. “You seemed to like…” makes an assumption and encourages an answer that doesn’t conflict. (Especially true for children).
Body language can influence whether someone will agree with a question.
Need to allow time to speak.
Probes used to get more information, especially neutral probes, eg. “Do you want to say anything else?”; or can prompt to help along if they’ve forgotten something.
Intended to be broadly replicable, so probing / prompting should aim to help along without introducing bias.
Focus groups.
Usually 3 - 10 (representative) people led by a trained facilitator.
In requirements activities it’s quite common to hold a focus group in order to identify conflicts in terminology or expectations from different sections in one department or organisation.
Benefit of focus groups.
Allows diverse or sensitive issues to be raised that might otherwise be missed. Assumes people develop opinions within a social context by talking to others.
Some points about conducting focus groups.
Question may seem simple, but the aim is for people to put forward their own opinions in a supportive environment.
Agenda developed to guide, but flexibility to follow unanticipated issues.
Facilitator guides / prompts, tries to get quiet ones to participate and prevent others from dominating.
Usually recorded for later analysis and people asked to explain comments further.
Planning and conducting interviews. (5 sections).
Developing questions / topics.
Collating documents to give interviewee (consent form, project description, etc).
Check / test equipment.
Work out structure.
Organise time / place.
Developing open / closed questions.
Open - goal of session exploratory.
Closed - need to know possible answers in advance; ideally ‘other’ not often used as an option.
Robson (2002) - guidelines for developing interview questions. (3)
Compound questions are confusing - split into separate ones.
Easier to respond and record.
People may not understand jargon or technical language and might not admit it, so explain in layman’s terms.
Try to ask neutral questions. Eg. not “why do you like…” - assumes they do like it and will discourage genuine answer.
Before the interview.
Make sure aims of the interview have been communicated to, and understood by, the interviewee, and that they feel comfortable.
Eg. finding about their world so you can dress, act and speak in a familiar manner.
Robson (2002) - Running the interview.
Better to listen more than to talk, respond with sympathy but without bias.
- Intro - explain purpose, reassure about ethical issues, ask if ok to record. Same for each.
- Warm-up session - easy, non-threatening questions first.
- Main session - logical sequence, more probing questions at the end.
Semi-structured - order may vary according to natural course. - Cool-off period - a few easy questions (defuse tension).
- Closing session - thanks and switch of equipment to mark end.
Other forms of interview. (2)
Telephone - if you can’t meet up. Like face-to-face but without the body language.
Online - either asynchronous (eg. email) or synchronous (eg. instant messaging).
May be preferable for sensitive issues to be anonymous.
Enriching the interview experience.
As often in neutral environment (away from desk) and is an artificial situation (away from normal tasks), it can be difficult to give full answers.
To help - props, eg. prototypes or work artifacts, or descriptions of common tasks.
These can provide a context for interviewees and help to ground data in a real setting.
Eg. keep a diary which questions will be based around.
Groupthink?
A phenomenon in which individual opinions become subsumed into that of the group;
a dominant member unduly influences the group.
Questionnaires. (3 points)
Clearly worded questions important when no researcher is present to encourage and resolve ambiguities.
Well-designed questionnaires are good at getting answers to specific questions from a large group of people, especially if spread across a wide area so infeasible to visit all.
Can combine with other methods, eg. information from a small number of interviews might be corroborated by sending questionnaires to a wider group.
Choice between questionnaires and structured interviews.
Similar questions - choice depends on motivation of respondent to answer:
- If high enough to answer without anyone else present, a questionnaire is cheaper and easier.
- If they need some persuasion, better to use face-to-face.
Eg. structured interview is easier and quicker where people won’t stop to complete a questionnaire, eg. train station, walking to a meeting.
Telephone interview lies between the two.
Developing questions for a questionnaire.
Can be harder than for interviews - no-one available to explain ambiguities.
Important questions are specific; where possible use closed questions with a range of answers (inc. ‘no opinion’, ‘none of these’).
Negative questions can be confusing and can lead to false information, but sometimes a mix of -ve and +ve questions can help check users’ intentions.
Questionnaire structure.
Many start with basic demographic information and details of relevant experience. Useful for putting responses into context, eg. web experience - different perspective may be due to experience level.
(Only relevant contextual info needed).
Specific questions that contribute to data gathering goals usually follow.
If long, may be subdivided into related topics so easier and more logical to complete.
General advice checklist for questionnaire design. (4)
Think about question order - the impact of a question can be influenced by question order.
Consider if different versions are needed for different populations.
Provide clear instructions on how to complete. Eg. say if you want a check put in one box.
Balance between white space and keeping compact. Long questionnaires cost more and deter participation / completion.
