Research Methodologies

Research Methodologies

Guohua Bai (Ph.D.) Blekinge Institute of Technology [email protected]

Introduction

- What is about research



   

2

The good researcher is not ”one who knows the right answers” but ” one who is struggling to find out what the right questions might be”. ”Original investigation undertaken in order to gain knowledge and understanding”. “It is all about asking the right question and then work systematically to address it.” Evaluation or change oriented research? Real world (field) or basic research?

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction - Academia vs. Industry: 

We are primarily concerned with research projects conducted in an academic setting. This means: – – –

3

that you should adhere to certain criteria regarding research methods and so forth. Academia: critical thinking, justification and develop your own thoughts, arguments, ideas and concepts. Industry: more oriented towards developing a particular solution to a problem. In particular, you must balance these two views if doing your thesis in industry.

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction

- Types of Methodologies

– – –

4

Fixed design: pre-specifications, assumptions, quantitative often Flexible design: qualitative often, unfold type Mixed: Use both in different sub-problems.

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction

- A typology of methodologies

5

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction

- Being scientific A

scientific attitude: carry out research systematically, sceptically, ethically  The standard view (positivism) of science p.20 box 2.1 and critiques p.22 box 2.2

6

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction - Some of the most common methods  Action

research  Experiment  Case study  Survey Those methods will be mainly introduced in this course.

7

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction - Example of research process (Blaxter. L.)

8

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction - Example of research process (Blaxter. L.)

9

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction

- process of doing research (Bai)

ort rep and ite /wr tion Re enta s n pre luatio eva

& se s, n aly An these clusio Syn w con Dra

, ory the ta se dy /ca a Stu lect d n test l Co desig and

lan e p g Tim rkin o e or w edur c pro

Guohua Bai (Ph.D. Docent) 08-10-23

10

s, tion ues oal m q is,g For othes dary hyp boun and

, ect ubj as and es ose Ch lems ratur b e pro rch lit sea

Motivation to do research

Introduction

- process of doing research 1. 2. 3. 4. 5.

6.

11

Choose a practically and theoretically meaningful subject Form your hypothesis/questions (see hypothesis) and predict possible outcome, i.e., the goal or contribution of your work. Work out a time table or procedure for the work (when, where to do what). Collect data, design experiment, and test hypothesis (see experiment design) Explain, analyse, or synthesise the result of your data or testing according to your hypothesis, and draw your conclusion Write report, evaluation

08-10-23

Guohua Bai (Ph.D. Docent)

Introduction - Inductive & deductive research process (Blaxter.L)

12

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design Design is concerned with turning research questions into projects This part deals with: – – – – – –

13

How to identify topics Originality, Types of research project, Tactics in project conduting, Hypothesis, Problem formulation.

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design - A general framework Theory

Purpose(s) Research questions

Data collecting

Methods

Fixed design strategy: The above are pre-defined, typical like an traditional experiment, normally involving hypothesis testing. Flexible design strategy: The above are evolved through the study or in the end of study, typical like ethnographic study, and even some case studies.

14

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design - Originality 

Think about how the ways of originality/innovation – – – –



Two original things you may do: – –

15

Tools, techniques, procedures and methods Exploring the unknown (rare) Exploring the unanticipated The use of data You can be original in the way you do things. You can be original by developing or producing something new

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design - Some hints 

Project diary! Start writing notes when your project starts and write continuously throughout the project (this is an important document for the course assignment and examination).





16

Divide the total time in the project among different tasks (how large should different tasks be relatively each other, relative time). This is actually a good piece of advice also when it comes to working with different assignments in courses in general. 08-10-23

Guohua Bai (Ph.D. Docent)

Research Design

- Choose a right subject  Your

knowledge about the subject  Your interest (your motivation)  Scientific and current interest  Challenge v.s. your capability (‘lagom’ in Swedish)  Empirical evidence available (case available)

17

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design - Risks to be not scientific  

 

Too general, or too specific (figure below) Too much raw work (e.g., coding, practical implementation) Descriptive (survey), not analytical (case study) Only reference to others’ work, no critical thinking 

Width



Too bad

Too general



18

Deepness

   

Problem domain Proper position

08-10-23 Too specific

Guohua Bai (Ph.D. Docent)

Research Design

- Decide types of projects  Five – – – – –

 The

19

types: Research-based project Development project Evaluation project Industry-based project Problem solving

five types are not mutually exclusive.

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design

- Estimate your project        

20

Capability to do it in time Interesting project Serious purpose (related to impact) Clear outcome Project suitably linked to your degree Sufficient scope and quality Not a personal issue Make a risk assessment 08-10-23

Guohua Bai (Ph.D. Docent)

Research Design - Finding a project  Several – – – – – –

21

ways of identifying a project:

List of projects (primarily industry) Past projects Talking with colleagues Reading around Clustering (keyword-driven) Brainstorming

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design

- Formulate hypothesis 

     

22

Hypothesis can be generalised in a semantic form: If [conditions] then [result] (not necessarily the same syntax form) Hypothesis should tell exactly what you are going to examine about (what is input, output, and their relationship). Hypothesis must be a meaningful statement (påstående) which is generally not answered yet. Hypothesis examples: If users are provided with a structured interface, then they will learn quicker than with an unstructured interface. Large organizations mostly employ recognized standard design methods for maintaining software quality. Probability of accident is higher when people talk in portable telephone under car driving 08-10-23

Guohua Bai (Ph.D. Docent)

Research Design

- From hypothesis to questions  Pick-up

key concepts in the hypothesis and ask youself ’is this concept theoretical and meaningful to study’  How the concepts related to each other  How to prove the assumed/hypothetical relationship  Then formulate questions around the concepts and relationship. 23

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design - Examples: hypothesis to questions 

  

Hypothesis 1: If users are provided with a structured interface, then they will learn quicker than with an unstructured interface. Key concepts: structured interace & unstructured interface, learning, and speed of learning (quicker) Relationship: more structured leads to quick learning Questions: how to categorize levels of interface structure? How users (novice & expert) learn to use a system through an interface? How to measure the speed of users’ learning?

Do the same excise as above to hypothesis 2 and 3.

24

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design

- Preparing a project proposal Mind-road for the proposal  Introduction

to the subject area (context)  Current research in the field (position where you are)  Identify a gap (problems identification, goals)  Identify how your work fills that gap (innovation, contributions, and method to implement a solution). 25

08-10-23

Guohua Bai (Ph.D. Docent)

Research Design - Project planning  This

is among the first things you have to do when the project has been approved.  Work breakdown  Time estimates (lead time and effort)  Milestone identification  Activity sequencing  Scheduling  Re-planning 26

08-10-23

Guohua Bai (Ph.D. Docent)

Managing your project  This – – – – – –

27

part will deal with

Managing time Mapping your project (Scheduling into time slots) Piloting (testing in small scale) Dealing with key actors Labor division Managing your temper

08-10-23

Guohua Bai (Ph.D. Docent)

Managing your project - Managing your time  Your

habits toward time  Time is short without motivation  Your priority dealing with parallel tasks  Time buffer to prepare unexpected happenings  By using project diary

28

08-10-23

Guohua Bai (Ph.D. Docent)

Managing your project

- Mapping your project (see also ex. p121, p122, Blaxter)

29

Week

Date

Deliverable

4

02/10

Created a schedule for our project.

5

02/17

Developed project idea. We presented our initial idea based on collaborative applications that featured shared displays.

6

02/24

Further refinement to the ideas occurred. Our project concept became focused on collaborative data collection and visualization. In studio, we presented our refined ideas.

7

03/02

First low-fi prototype was shown to class. Revisions were suggetested.

8

03/09

First revision of Low-Fi prototype will be presented. User testing will take place this week.

9

03/16

user testing will continue and rapid prototype will begin to be developed

10

03/23

Spring Break

11

04/06

Testing and development of rapid prototype..

12

04/13

Evaluation of rapid prototype.

13

04/20

Development of final Interactive Prototype: First Half

14

04/27

Final Interactive Prototype: Second Half will be developed.

15

05/04

Pilot User Studies will occur.

16

05/08

Final Iteration on Prototype will occur. Final reserach product will be released.

17

05/15

Presentation of project.

08-10-23

Guohua Bai (Ph.D. Docent)

Managing your project - Piloting  Never

believe things will go according to the planed.  Strategy to select the parts of project for pilot.

30

08-10-23

Guohua Bai (Ph.D. Docent)

Managing your project - Dealing with key actors  Supervisor:

expectation & self-responsibility  Project managers: knowing resources for projects  Examiner: knowing criteria of examination  Project owner: making a contract with those who have right to decide access to data sources, and financing 31

08-10-23

Guohua Bai (Ph.D. Docent)

Managing your project - Labor division  Team

work and co-operation must be based on a clear labor division.  Shared responsibility  Positive attitude toward different opinions (contradictions as source of development, avoid conflicting)

32

08-10-23

Guohua Bai (Ph.D. Docent)

Managing your project - Managing your temper 

 



33

Remeber research is a process of learning and you are allowed to make mistake. 20 often made mistakes (p136, box 66, Blaxter) Expect changes and mistak and never loss your temper Prepare youself according to those 20 mistakes in box 66, what to do if you made the mistake?

08-10-23

Guohua Bai (Ph.D. Docent)

Data Collection and Analysis  This – –

part deals with the followings:

Data- what is about? Techniques for collecting data, including:  Reading

documents  Survey and Questionnaries  Observation  Interview –

34

Quantitative and Qualitative analysis

08-10-23

Guohua Bai (Ph.D. Docent)

Data Collection and Analysis - What is about data?  

 

35

Data – Information – Knowledge – Wisdom Information is a form of processed data or facts about objects, events or persons, which are meaningful for the receiver, inasmuch as an increase in knowledge reduces uncertainty. Information becomes knowledge only when we decide to put it into use. Information I= i (D,S.t), where i represents an interpretation process, D a set of data, S: previous knowledge, and t: the time which the information is available. 08-10-23

Guohua Bai (Ph.D. Docent)

Data Collection and Analysis - Reading documents  Regarded as secondary and context of the documents) – – – – – –



36

data (knowing the author

Statistics and report Government white papers Company annual report Institutional documents Books and journals Newspaper, television and radio

Why using secondary data? (p153 box 74 ) 08-10-23

Guohua Bai (Ph.D. Docent)

Data Collection and Analysis - Surverys and questionnaries  Steps

in conducting interview-based questionnare survey （ 100% time). – – – – –

37

Develop research questions, and study design (25%) Informal test of questionare and revise questionnaire (5%) Carry out main data collection interview (30%) Code data and data files (15%) Analyse data and write report (25%) 08-10-23

Guohua Bai (Ph.D. Docent)

Data Collection and Analysis - Surveys and questionnaries A

survey can be done by face to face interview (time consuming), telephone, postal reply (low response), and increasing via Internet or e-mail (representative validity?)

38

08-10-23

Guohua Bai (Ph.D. Docent)

Data Collection and Analysis - Observation  Participation

in observation (action research)  Independent observation  Structured observation and open mind observation  Observation in enthnography study  Use tools, dictaphone, video, etc. to observation 39

08-10-23

Guohua Bai (Ph.D. Docent)

Data Collection and Analysis - Interview  The

most powerful way to collect data  Structured & open-ended interview  Select right places to conduct interview  Work in team, one conducts questioning, another takes notes  Let the interviewees prepare questions  Think about how to deal with the data collected 40

08-10-23

Guohua Bai (Ph.D. Docent)

Quantitative and Qualitative Analysis  Quality

& quantity-two sides of the same

coin?  Quality first (knowing what) and quantity later (knowing how much)  The dialetical and independent relationship between quality and quantity

41

08-10-23

Guohua Bai (Ph.D. Docent)

Quantitative and Qualitative Analysis - Analysis     

42

Data analysis (interpretation)  meaning (information) Unstructured data analysis (representation) structured data Mass data  Analysis (Abstracting) visible and key findings Analysis as a break-down or ’drill-in’ process (not a summary or add-up process) Analysis must apply theories/concepts to link your findings to existed knowledge/understanding. 08-10-23

Guohua Bai (Ph.D. Docent)

Quantitative and Qualitative Analysis - Process  There

is no standard method to analyse all kinds of data (documents, interview, observation, questionaries) p.185-200, Blaxter)  The right choice of a method for data analysis depends on the quality of data, purpose of study,and your knowledge.  Using statistics software for analysis quantitative analysis 43

08-10-23

Guohua Bai (Ph.D. Docent)

Some Basic Types of Research Methodologies (* indicates the weight on the subject)

Methodologies Questions used data collection Survey who, what, where, how many(much)

Application

Contemporary theory events

Tools or instruments and

*

****

statistics, interview

Exploratory

what, where, how

**

***

participatory, ethnography

Experiment

What, how

**

**

prototyping, technique

Case Study

Why, how

****

***

Modelling

what, how

****

**

logic, mathematics

Theory generalisation

what, how, why

*

inductive, intuition, logic

44

*****

08-10-23

systems analysis, synthesis

Guohua Bai (Ph.D. Docent)

Basic Methods - Experiment 



45

Experiments are carefully planned and fully controlled. An experiment should be replicable, i.e. somebody else should be able to repeat it. You observe or measure outputs of the examined system (dependent variables) by systemically change/manipulate the value of the input (independent variables), and to identify if the relationship of input-output agrees with your hypothesis.

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Methods

- Experiment principle

46

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Methods - Experiment 

Example: to test if different rates of learning (output) exist for three different type of interfaces (input as command, ask/answer, window based), this examination can be shown as this figure: Environment (disturbance)

Input (independent variables) (types of interfaces)

Examined System or phenomena

Output (Dependent Variables) (Rate of Learning)

Hypothesis of In/out relation

47

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Methods

- Experiment process

48

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Methods

- Experiment planning

49

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Methods

- Experiment operation

50

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Methods - Case Study

What is case study? The essence of a case study, the central tendency among all types of case study, is that it tries to illuminate a decision or set of decisions: why they were taken, how they were implemented, and with what result. (Schramm, 1971, emphasis added)

51

–

We extend the topic ‘decision’ to include ‘individuals’, ‘organisations’, ‘process’, ‘program’, ‘methodologies’, ‘policies’, ‘events’, ‘society’ in our case studies.

–

Emphasis on explanatory, not descriptive, and on single- embedded (type 2) , not on multiple-embedded (type 4, see table 1.).

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Methods

- Basic types of case study Table 1: basic types of case study multiple-case

Singles-case Holistic (single unit of analysis) Embedded (multiple unit of analysis)

52

Type

1

Type 2 08-10-23

Type 3

Type 4

Guohua Bai (Ph.D. Docent)

Basic Methods (end L5)

- Choose a right case The case represents the critical case to confirm, challenge, or extend a theory by testing its proposition, or alternative explanations. ☛ The case represent an extreme or unique case. ☛ The case is an revelatory (avslöjand, uppenbarande) case.

Why (not) to choose multiple case study? Evidence is more compelling, robust, and representative. Replication logic (every case should serve a specific purpose within the overall scope of inquiry, not sampling logic (multiple respondents in a survey) Extensive resource and time, beyond a single student or researcher

53

08-10-23

Guohua Bai (Ph.D. Docent)

Measurement

54



1: Nominal Scale: A classification of categories, qualitative differences, no quantitative meaning (e.g., group 1, group 2,......).



2: Ordinal Scale: Ordering according to quantitative differences, though it does not exactly tell how much the differences are (good, bad, ..).



3: Interval Scale: Assigning an equidistant to different status of measured phenomena (e.g., Fahrenheit temperature scale)



4: Ratio Scale: There is an absolute zero point so we can make ratio calculation (e.g, distance, mass, time) 08-10-23

Guohua Bai (Ph.D. Docent)

Reliability (method), Validity (data), and generalisability (analyses) 





Reliability is judged by the consistency with which known method (instruments) produce certain measurement. If you measure the same phenomenon more than once with the same instrument, method, then you should get the same result (free of subjective-bias, standard methods) Validity is judged by whether you are measuring, or explaining what you claim to be measuring or explaining (Validity of data generation methods and validity of interpretation). Generalisability is judged by the extent which your explanations are still valid and reliable (deduction).

Accurate (reliable), and correct (valid) Accurate (reliable), not correct (not valid) Not accurate (not reliable), correct (valid)

55

08-10-23

Guohua Bai (Ph.D. Docent)

Criteria for a Valid Examination

56



1: The examiner must systematically manipulate/observe one or more independent variables in the domain under investigation;



2: The examination must be made under controlled conditions, such that all variables which could affect the outcome of the experiment are controlled;



3: The examiner must measure output as a function of the input variables.

08-10-23

Guohua Bai (Ph.D. Docent)

Conditions for a valid and reliable examination  





57

System being observable: To transform quality to quantity: System being stable: The system must be able to remain at some determined status under constant experimental conditions. System being controllable :The system output (dependent variables) must be sensitive to the change of input (independent variables). (The selected dependent variables and independent variables must have an observable cause-effect relationship. ) You must have a goal for an experiment, e.g, to test some predefined hypotheses, to gain more information about problem space, to identify variables in relation to a problem under investigation.

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Methods

- Prototyping in Design  To

understand, evaluate and validate a product solution, a prototype may be built. We may do the same for processes/methods using empirical studies.  Especially, experiments using human subjects or simulation may be used for process evaluation. 58

08-10-23

Guohua Bai (Ph.D. Docent)

❶ Lifecycle or Waterfall Approach

59

08-10-23

Guohua Bai (Ph.D. Docent)

Problems About the Iifecycle Model

1: Real projects are not sequential in the rigid way that this model assumes, but in a different order with iteration. 2: It is not possible to elicit or identify all the requirements at the start of the project because of unpredictable changes. 3: It is often expensive to correct design and coding errors in the late process of testing and maintenance.

60

08-10-23

Guohua Bai (Ph.D. Docent)

Advantage About the Lifecycle/Waterfall Model

1: It provides a comprehensive template (as of DNA) in which many important aspects of design can be placed. 2: It provides generic steps that are found in most software engineering paradigms. 3: It is the most widely used model at least in large software project. 4: It is superior than unplanned design (haphazard).

61

08-10-23

Guohua Bai (Ph.D. Docent)

❷ Prototyping Approach 

62

Introduction: The use of experimenting with prototypes has become prominent due to a number of claimed advantages in the case that users requirements are difficult to specify. In principle users should be highly motivated in acting since they are provided with more chances to improve their work, to verify if their needs are taken care of and that the terms used in the interface, functions of the designed system are consistent with their work. 08-10-23

Guohua Bai (Ph.D. Docent)

Types of Prototypes









63

Throwaway prototype: to initiate user interest, develop builder skills, to reduce risk and investment. Evolving prototype: Adaptive, prototype to be product Co-operative Prototype: Participatory Design (PD) Embryonic Prototype: Feedback Learning, organic development. 08-10-23

Guohua Bai (Ph.D. Docent)

Model of Prototyping Approach

64

08-10-23

Guohua Bai (Ph.D. Docent)

Steps in Prototyping 

 







65

Requirements Gathering: Designers and the customer define the overall objectives of the system, and some known requirements. Quick Design: Focuses on the design of interface. Build Prototype: Choose the types of prototypes and build quickly. Evaluate and Refine: Involving the designers, the customers, and the users to experience the prototype and elicit requirement more specific than requirement gathering. Engineer product: Based on the prototype, the final product is engineered. Iteration: Occurs at any stage where an error is detected or an evaluation is identified, until the product is satisfied by the users and customers. 08-10-23

Guohua Bai (Ph.D. Docent)

Dimensions of Prototyping

66

08-10-23

Guohua Bai (Ph.D. Docent)

Techniques for Implementing Prototypes      

67

Verbal prototyping: verbal description of choices and results. Paper mock-ups: printouts or sketches of screen designs. Wizard of Oz: human expert operating behind scenes. Fake data: similar data, images instead of video, etc. Simple algorithms: ignore special cases. Prototyping tools: e.g. HyperCard, ToolBook. UIMS (User Interface Management Systems): interactive interface builders such as Visual C++. 08-10-23

Guohua Bai (Ph.D. Docent)

EmPA System Development Model Weeks or Monthes

User satisfied

Architectural Kernel Days or weeks

Days or weeks

(Re)design Prototype Start with

Work with prototpye

Hours or days

Work done

Work Evaluation

Communication Links User not satisfied Hours or days

Feedback to designer

68

08-10-23

Guohua Bai (Ph.D. Docent)

Problems with Prototyping Approach

☛ "Throwaway" type of prototype will have adaptive problem in case the customer or user will hold and work with the prototype ☛ It is a trial and error approach to design, so it could be expensive in view of customers time and resources. ☛ It is a learning-oriented and feedback communication process (participatory Design), there will be many contradictions involved the whole process. ☛ Inadequate problem analysis. ☛ Difficulty in resource planning.

69

08-10-23

Guohua Bai (Ph.D. Docent)

Problems with Prototyping Approach

☛ difficulty in making a good decision on whether to enhance an old system or build a new version

☛ the boredom that the Nth iteration of a system may bring to the designers one iteration after another

☛ difficulty for documentation, because the system is so easily changed that keeping documentation up to date is a problem.

☛ Difficulty to develop large system because of multi-site testing and difficulty in integration

70

08-10-23

Guohua Bai (Ph.D. Docent)

Advantages with Prototyping Approach

1. The customers and users can be motivated in acting since they are provided with more chances to improve their work, to verify their needs, to check if the interface, functions are consistent with their work. 2. Errors in design can be checked out in an early stage.

71

3. The final product can be more suitable and accepted by the specific users than lifecycle model. 08-10-23 Guohua Bai (Ph.D. Docent)

When Not to Use Prototyping

◆ ◆ ◆ ◆

72

when the designers do not have a good understanding of the programming tools when the organisation's data resources are not organised and managed for quick access when information department is unwilling to develop a staff of professional prototype builders the users is unwilling to invest time during the development

08-10-23

Guohua Bai (Ph.D. Docent)

How Prototyping differs

Different tools ◆ Different Skills: ◆

◆ ◆ ◆

73

Interview skills are not as important as they are with conventional methods Designers spend more time with users and less time coding (more people oriented) On user’s side, prototyping requires the most knowledgeable users (the manager) and more time to work on the system design. 08-10-23

Guohua Bai (Ph.D. Docent)

❸ Participatory Design Approach

PD

The Scandinavian tradition of sociotechnical approach to the design of work and artefacts 74

08-10-23

Guohua Bai (Ph.D. Docent)

Two objectives of PD  Increase  Increase

75

the satisfaction of the end-users

their work efficiency, effectiveness, and aesthetics

08-10-23

Guohua Bai (Ph.D. Docent)

Three levels of participation  Consultative participation: leaves the bulk of decisions on how a new work system shall be designed and jobs structured to the traditional systems design group.  Representative participation: requires a higher level of involvement from the staff of a user department.  consensus participation: takes the democratic approach to a higher level by attempting to involve all staff in the user department continuously throughout the system design process.

76

08-10-23

Guohua Bai (Ph.D. Docent)

PD Condition and Conflict

☛

☛

77

Conditions for PD approach: No-one must suffer and as many employees as possible should gain from the changes. All change involves some conflicts of interest. To be resolved, these conflicts need to be recognised, brought out into the open, negotiated and a solution arrived at which largely meet the interest of all parties in the solution. PD approach is one way of bringing about differences and conflicts of interest into the open.

08-10-23

Guohua Bai (Ph.D. Docent)

What to take care of when the project is established ☛ ☛

☛ ☛

78

Have Management support Specify in a contract how much time the users can/shall spend on the project. Have a steering group in which conflicts can be discussed. Be sure the required equipment is available for systems experts and users.

08-10-23

Guohua Bai (Ph.D. Docent)

What to take care of during the project

79

☛ If the users don´t want some thing, don´t do it, even if it seems to be the perfect computer systems for their needs from a systems expert point of view. ☛ Do not do everything the users proposed. The users are not always aware of the consequences of their proposals, and they can often be conservative in that they do not utilise the possibilities given by computer technologies. ☛ Do not forget to inform users about any progress, otherwise the users will feel left out, and it will no longer be a PD project. 08-10-23

Guohua Bai (Ph.D. Docent)

Problems in PD

 PD

is too idealistic;  PD is biased toward workers;  PD lack method or model;  PD designers need to rely strictly on experience

80

08-10-23

Guohua Bai (Ph.D. Docent)

Prototyping in PD  Prototyping

as learning process between designers and end-users  prototyping as skill learning for both endusers and designers  Problems in PD prototyping.

81

08-10-23

Guohua Bai (Ph.D. Docent)

System Thinking and Methodologies •What is about GST –A theory of everything (interdisciplinary, cross-scientific, universal, law of laws)? –Enlightenment v.s. instrument (why, what and how) –Abstract v.s. Specialised

•Words by Kenneth Boulding (1956): General Systems Theory is the skeleton of science in the sense that it aims to provide a framework or structure of systems on which to hang the flesh and blood of particular disciplines and particular subject´matters in an orderly and coherent corpus of knowledge.

82

08-10-23

Guohua Bai (Ph.D. Docent)

System Thinking and Methodologies Why Study System Methodologies • • •

Introduce order and meaning to observations that may otherwise seem chaotic. Explain complexity in a simple manner by establishing connections and laws. Guide future research direction, prediction, make recommendations, and find problem solving methods.

The stages of human knowledge in the history Pre-science (Scholastic paradigm) (The theological stage) 2 Century

83

Rise of Science (Renaissance paradigm) (metaphysical stage) 16 Century

Mechanistic and determinism; hegemony of determinism; Positive Stage 18 century

08-10-23

2000 Relativity and quantum theory 19 century

The systems age Mid. 20 century

Guohua Bai (Ph.D. Docent)

The Reductionism Thinking and Analysis Reductionism argues that from scientific theories which explain phenomena on one level, explanations for a higher level can be deducted. Reality and our experience can be reduced to a number of indivisible basic elements. Qualitative properties are possible to reduce to quantitative ones The reductionism provide foundation for the analytical method : Dissect conceptually/physically Learn the properties/behaviour of the separate parts From the properties of the parts, deduce the properties/behaviour of the whole. Free from environment influence: standard scientific laboratory Free from observer’s subjective influence: non-intervention, neutrality and objectivity

84

08-10-23

Guohua Bai (Ph.D. Docent)

The Scientific Method In conducting scientific activities (such as to describe, control, predict, and explain phenomena) was formed by the following orders:     

85

Reduction of complexity through analysis Development of hypotheses Design and replication of experiments Deduction of results and rejection of hypotheses Evaluation and correction

08-10-23

Guohua Bai (Ph.D. Docent)

What is the big thing ? - A metaphor of analysis

86

08-10-23

Guohua Bai (Ph.D. Docent)

Some propositions from systems scientists "The whole is more than the sum of its parts" (Aristotle, 384-322 B.C. Greek) is a definition of the basic system problem. "The properties and model of action of higher levels are not explicable by the summation of the properties and modes of action of their components taken in isolation. If, however, we ensemble of the components and relations existing between them, then the higher levels are derivable from the components" That is, in order to understand an organised whole we must know both the parts and the relations between them .

87

(Ludwig Von Bertalanffy, 1972)

08-10-23

Guohua Bai (Ph.D. Docent)

Systems Synthesis Synthesis is a prerequisite for the systems thinking, just as analysis was for the mechanistic thinking. It takes steps from whole to parts: 1. Identify the system of which the unit in focus is a part; 2. Explain the properties or behaviour of the system; 3. Finally, explain the properties or behaviour of the unit in focus as a part or function of the system. Synthesis does not create detailed knowledge of a system’s structure. Instead it creates knowledge of its function. Therefore, synthesis must be considered as explaining (why, functions, understanding) while the scientific method must be considered as describing (what, how). @

88

08-10-23

Guohua Bai (Ph.D. Docent)

Th

eb

ig

ec o

log

ys

ys t em

Why does the big thing behaviour as such ? - A metaphor of synthesis

89

08-10-23

Guohua Bai (Ph.D. Docent)

Basic Ideas of System Thinking What is a “system”

90

☛

A system is not something presented to the observer, it is something to be recognised by him. Most often the word does not refer to existing things in the real world but rather to a way of organising our thoughts about the real world (Constructivist view)”

☛

”A system is anything that is not chaos. (Kenneth Boulding, 1985)”

☛

”A system is a structure that has organised components (Churchman, 1979)

☛

A system is a set of interacting units or elements that form an integrated whole intended to perform some function (Lars Skyttner, 1996).

☛

A system is the organised collection of men, machines and material required to accomplish a specific purpose and tied together by communication links (Business management). 08-10-23

Guohua Bai (Ph.D. Docent)

System A relationship An element

ut

Outp

Inpu t

’The system’ ’The environment’ These three elements are on a feedback loop

91

Boundary

A general conception of ’system’ (Flood R. and Jackson M., 1991) 08-10-23

Guohua Bai (Ph.D. Docent)

Kinds of Systems ☛

(Ludwig Von Bertalanffy,1972)

Real System (concrete, or physical) which entities are perceived in or inferred from observation and existing independently of an observer; they can be man-made or natural, living or non-living systems

☛

Conceptual Systems which essentially are symbolic idea constructs(linguistic, mathematics, logic);

☛

Abstracted System which is conceptual systems corresponding with reality (model of traffic, a bridge)

☛

92

Living and Organic Systems: see p55 specific qualities differentiate living systems from non-living ones (Structure & genetic; co-ordination &subordination, control &freedom; qualitatively transform &continuously renew). 08-10-23 Guohua Bai (Ph.D. Docent)

Kinds of Systems ☛

(Ludwig Von Bertalanffy,1972)

Open/closed systems

An open system (all living systems) is always dependent upon an environment with which it can exchange matter, energy and information, while a closed system is only open for the input of information (energy). The differences between open and closed systems are relative. An organism is a typical example of open system but, taken together with its environment, it may be considered as a closed system. Closed systems is not isolated systems which are totally closed from their environment, and the concept is not applicable in reality (cosmos is environmentless, and hence it is isolated) Environment (not the boundary) should be seen as those parts that 1) beyond the direct control the system, and 2) yet they influence the behaviour of the system. The immediate environment is next higher level of system minus the system itself. The entire environment to a system includes this immediate environment plus all systems at higher levels which contain it.

93

08-10-23

Guohua Bai (Ph.D. Docent)

Properties of Systems ☛

Interrelationship and Interdependence of objects and their attributes Unrelated and independent elements can never constitute a system.

☛

Holism Holistic properties not possible to detect by analysis should be possible to define in the system.

☛

Goal seeking Systemic interaction must result in some goal or final state to be reached or some equilibrium point being approached.

☛

Transformation process All systems, if they are to attain their goal, must transform inputs into outputs. In living systems this transformation is mainly of a cyclical nature.

☛

Inputs and outputs In a closed system the inputs are determined once and for all; in an open system additional inputs are admitted from its environment.

94

08-10-23

Guohua Bai (Ph.D. Docent)

Properties of Systems ☛

Entropy This is the amount of disorder or randomness present in any system. All non-living systems tend to toward disorder; left alone they will eventually lose all motion and degenerate into an inert mass. When this permanent stage is reached and no events occur, maximum entropy is attained. A living system can, for a finite time, avert this unalterable process by importing energy from its environment. It is then said to create negentropy, something which is characteristic of all kinds of life.

☛

Regulation The interrelated objects constituting the system must be regulated in some fashion so that goals can be realised. Regulation implies that necessary deviations will be detected and corrected. Feedback is therefore a requisite of effective control. Typical of surviving open systems is a stable state of dynamic equilibrium.

95

08-10-23

Guohua Bai (Ph.D. Docent)

Properties of Systems ☛

Hierarchy Systems are generally complex wholes made up of smaller subsystems. This nesting of systems within other systems is what is implied by hierarchy.

☛

Differentiation In complex systems, specialised units perform specialised functions. This is a characteristic of all complex systems and may also be called specialisation or division of labour.

☛

Equifinality and multifinality A profound difference between most closed systems and open systems is expressed by the concept of "equifinality." In an closed system the final state of the system is determined by its initial conditions. A change in the initial conditions produces a change in the final conditions. A different behaviour is shown among open systems : under many conditions the same final state may be reached from different ways.

96

08-10-23

Guohua Bai (Ph.D. Docent)

Example of Systems Hierarchy

97

Holon Hierarchy represented by circles 08-10-23

Guohua Bai (Ph.D. Docent)

Hierarchy of Systems

98

Multilevel of systems hierarchy

An organism hierarchy 08-10-23

Guohua Bai (Ph.D. Docent)

Checkland’s Soft Systems Methodology

  



Step 1: The problem situation: unstructured Step 2: The problem situation: expressed Step 3: Root definitions of relevant systems (CATWOE) Step 4: Conceptual model Step 5: Comparison of 4 with 2 Step 6: Feasible, desirable changes Step 7: Action to improve the problem situation



Critique of SSM

     

99

Philosophy: hard and soft paradigms Principles: Learning; Participation; Culture. Methodology (7 steps)

08-10-23

Guohua Bai (Ph.D. Docent)

Checkland’s Soft Systems Methodology

The problem situation: unstructured 1

Action to improve the problem situation

7

2 The problem situation: expressed

5

Feasible, desirable 6 changes: -----------------

Real World

Comparison of 4 with 2

Systems thinking 4

Conceptual models 3

Root definition of relevant systems

4a Formal system concept

4b Other systems thinks

Activity Theory Approach What is an Activity

101

–

Activity is a necessity of human social life

–

Activity is an interaction between subjects and their physical, social and cultural environment

–

Activity is a reciprocal transformation between subject and object.

–

Activity can be analysed in a hierarchical structure.

–

Activity is always mediated by artefacts.

–

Activity is initially social in nature.

–

Activity theory considers contradictions as a basic resource of development.

08-10-23

Guohua Bai (Ph.D. Docent)

Activity Theory Perspective

Design  What

is design?…It’s where you stand with a foot in two worlds — the world of technology and the world of people and human purposes —and you try to bring the two together’

Mitch Kapor in T. Winograd Bringing design to Software (1996), p.1)

103

08-10-23

Guohua Bai (Ph.D. Docent)

Interaction – A feedback loop Feedback

By using tools

Subject

Acting on

Object

Actors

104

Transform

Being acted

08-10-23

Outcome Result

Guohua Bai (Ph.D. Docent)

IT Design – More Than an Actions Activity/Motives Why? For what purpose?

S

Action: Design IT

Outcome

analysis, design, implementation, test

Operations coding software, running test, etc.

105

08-10-23

Guohua Bai (Ph.D. Docent)

Tool Mediation (Leontiev) Tool

Activity Actions Operations

Motive Goals Tasks Object

Subject

106

08-10-23

Outcome

Guohua Bai (Ph.D. Docent)

Activity Structure: the Triangle (Y. Engestrom) Tools

Transformatio n Object process Outcome

Subject

Rules

107

Community 08-10-23

Division of labour Guohua Bai (Ph.D. Docent)

Learning Activity in IT Design Feedback-learning

Hard/software Designers’ outcome Designers

Contracts, Agreement Time table, standards

IS in Design

IS Companies, research units

Designers’ Activity System

108

IS in Use

Dept. of Analysis,Coding, Evaluation, User Education

Rules, norms, standards, etc. for work

08-10-23

Users

Work

Users’ work communities, other organisations

Users’ outcome

Work distribution and responsibility

Users’ Activity System Guohua Bai (Ph.D. Docent)

109

08-10-23

Guohua Bai (Ph.D. Docent)

The simplified IMIS architecture IMIS Web Service

Medical Care Hospital

Patient home or out Other services

Other services providers

Home care and service

Home services Man-materials flows Information flows

110

08-10-23

Guohua Bai (Ph.D. Docent)

IMIS project context Component 3: Web based Intelligent monitoring and alarming system

Hospital

Web Service

Home services

Elderly home

Component 2 (data collection/fusion and transition, gateway)

Component 1 (needed devices with sensors) Other services providers

111

Component 4 A test environment (Lab.) 08-10-23

Guohua Bai (Ph.D. Docent)

Practicing the triangle in e-health Eva’s Working tools

Eva’s own info. Eva’s care Receivers

Eva’s IMIS Platform

(Eva) Healthcare Operations in the field

Personal rules Eva’s Virtual Community

112

08-10-23

Eva’s Outcome

Eva’s taska and responsibility

Guohua Bai (Ph.D. Docent)

AT as a construction tool for IT architecture Evaluation and measures

Team’s working tools

Teams’ info.

IMIS Web services

Healthcare Receivers

Healthcare (team) outcome

Team leader Rules, agreements in Healthcare,

Healthcare labor division

To field workers Healthcare Community

113

08-10-23

Guohua Bai (Ph.D. Docent)

Networked and distributed activity in healthcare ld e fie

th rom f n o

Portable

i luat Eva

Information system

Field worker 1

Healthcare outcome (field 1)

Field 1 E-health warehouse

Other field nodes Healthcare (team) outcome

Team leader 1

Healthcare outcome (field 2)

Ev alu

atio

114

nf rom

Portable Information system

the

fie

ld

Field worker 2

Field n

08-10-23

Guohua Bai (Ph.D. Docent)

An example of IMIS interface

115

08-10-23

Guohua Bai (Ph.D. Docent)

What does AT help us to understand? – – – –

116

A broader view of the IT Understanding IT as a tool for work Structured observation of the “real life situations” (social and cultural context) Construction IT as info-logical world of real activity

08-10-23

Guohua Bai (Ph.D. Docent)

What is about IMIS 

The IMIS project is to build up a shared, mobile, and web based communication platform for both diabetic healthcare providers (hospitals, nurses, home services, family members etc.) and the diabetic patients to communicate each other through some intelligent terminals (such as IBG: Intelligent Blood Glucose, IM: Intelligent Monitoring, II-Pen: Intelligent Insulin-Pen, VR: Vibrating receiver Device) as shown in figure (next page)

117

08-10-23

Guohua Bai (Ph.D. Docent)

Cybernetics 

In 1948 Norbert Wiener defined the ”Cybernetics” in his book Cybernetics or control and communication in the Animal and the Machine. Later on it became part of GST.



The words Cybernetics is derived from the Greek noun, Kubernetes, which associates to pilot or rudder.



The essential feature of intelligent machine and human is that they must operate according to feedback-the control of a machine or man on the basis of its actual performance.



The black-box approach (Input-processing-output -feedback) Input

118

Processing

Output

Feedback 08-10-23

Guohua Bai (Ph.D. Docent)

Basic Elements of Cybernetic Systems 

A Cybernetic system consists of five basic elements (which are themselves subsystems). They are: – – – – – –

A control object, or the variables to be controlled A detector, or scanning subsystem (sensor) A comparator An effector, or action-taking subsystem A transformer (Feedback) Input and output Input t

ara p m Co

119

+ or

Effector

Output

Object Detector Feedback Transformer

08-10-23

Guohua Bai (Ph.D. Docent)

Cybernetics (Example 1) - Control the speed of an engine

Feedback Regulation

James Watt’s speed controlling regulator Engine speed changes generates counteracting forces from the regulator. The steam is chokes or released, there by returning the engine to normal operating speed.

120

08-10-23

Guohua Bai (Ph.D. Docent)

Water entry

Float Open Closed

121

08-10-23

Guohua Bai (Ph.D. Docent)

Cybernetics (Example 2) - Control the temperature of an oven

V(500)

Controlled Oven

E-Input set

(Goal)

Input-Output Compare H-Output

V(250)

E

H

E

H

(at 250

_ e) +

Feedback loop V(0)

Output Measurement

122

E : Electrical type signal H : Heat type signal Transforming to e : Allowed deviation 08-10-23

Guohua Bai (Ph.D. Docent)

Cybernetics (Example 3) - Control the navigation of a ship

Information System

Target

Human Operator

e(t)

Control System

F(t)

Executive system

U(t)

Cursor

Feedback

123

08-10-23

Guohua Bai (Ph.D. Docent)

O(t)

Cybernetics (Example 4) - Feedback Learning Input Receptor (Filter)

Learning and decision Unit +

Effector (Executive unit)

Output

_

Comparator

Goal Setter

A Learning System With a Positive and Negative Feedback

124

08-10-23

Guohua Bai (Ph.D. Docent)

Cybernetics (Example 5) - Management Cybernetics

θi

η

+

Operator 2

+

Operator 4

+

µ

Operator 1

θo

Operator 3

θi

: Input desired production goal θL : New orders arrived θo : Output of production

125

θL

Based on S. Beer (1959) Cybernetics and Management, New York: Wiley & Sons, PP171 08-10-23 Guohua Bai John (Ph.D. Docent)

Cybernetics (Example 6) - City Government Cybernetics

Disturbances Recommended objectives

PBS

+

Executive and legislative decision

Administration

City and its people

State of the city

_

MIS

126

Source: Adapted from E.S. Savas, ”City Hall and Cybernetics,” in Cybernetics and the management of Large systems, ed. Edmond M. Dewan, American Society for Cybernetics, New York: Spartan Books, 1969. pp134-135

08-10-23

Guohua Bai (Ph.D. Docent)

Example (Sociocybernetics) Open for Enquiring and Learning Military, political or democratic decisions Materials needs for re-production Social law, rules, norm, culture, etc,

-

Market/Planned economy control

Production

Distribution

(IS design)

(IS marketing for re-design)

Market needs for consumption Consumption (IS in use)

Exchange (IS marketing) Inner loop sociocybernetic system

Social sensor Statistics bureau Mass media Investigator Outer loop sociocybernetic system

127

08-10-23

Guohua Bai (Ph.D. Docent)

Cybernetics (Negative and Positive Feedback) Situation at the start

Equilibrium

Situation at the start Time

Negative feedback

Situation at the start

A system with positive feedback will eventually run away, explode and destroy the system.

Explosion

Blocking

128

A system with negative feedback will approach the goal - an equilibrium state no matter the initiate status, and automatically compensate for disturbing forces not necessarily known where it comes and what is the disturbing force.

Positive feedback can be used only temporarily. Time

Positive feedback 08-10-23

Guohua Bai (Ph.D. Docent)