Software Business and Engineering Institute

1

SoberITSoftware Business and Engineering Institute

HELSINKI UNIVERSITY OF TECHNOLOGY

S200110U-VI1(21)

Constructive Research

Timo Soininen, Casper Lassenius, Jari Vanhanen

2



S200110U-VI2(21)

Contents

What is constructive research?Why constructive research?Phases of constructive researchEvaluating constructive researchBuilding blocks for constructive research in Software EngineeringPhilosophical musingsExercise

3



S200110U-VI3(21)

What is Constructive Research?Constructive research

Aims at producing novel solutions to practically and theoretically relevant problemsManagerial problem solving through the construction of models, diagrams, plans, organizations, etc.Widely used in software engineering and computer science, rarely in management and social sciencesThe engineering research traditionOften involves other approaches (qual. & quant.)A.k.a. design research,

ConstructionAn entity, which produces a solution to an explicit problem

mathematical algorithmMorse alphabetactivity-based costing (ABC)

4



S200110U-VI4(21)

Shaw, 2001

5



S200110U-VI5(21)

Shaw, 2001

6



S200110U-VI6(21)

Hevner et al. 2004

7



S200110U-VI7(21)

Hevner et al. 2004

8



S200110U-VI8(21)

The Constructive Approach as a Methodology

Is a type of applied studiesproduction of new knowledge in the form of normative applicationscreates a new reality – does not try to understand, explain, classify, etc. the existing one

Constructive approach vs. basic studies have no explicit normative purposesdevelopment of techniques purely aim at improving skills and meansanalytic model building (applied studies) has unclear practical adequacyscientific problem solving may produce unique solutionsconsulting does not presuppose use of scientific methods

9



S200110U-VI9(21)

Phases of the Constructive Research Process

Idealized model1. Find a practically relevant problem2. Obtain an understanding of the topic and the problem3. Innovate, i.e., construct a solution idea

heuristic processtheoretical justification and testing come later

4. Demonstrate that the solution works5. Show theoretical connections and research contribution6. Examine the scope of applicability

In practice the steps do not follow each other in a simple sequence - the process is both iterative and sometimes recursiveQ: How long does this take?

10



S200110U-VI10(21)

1. Finding a relevant problem2. Preunderstanding

Finding the problemSources of problems

LiteratureColleaguesOwn experience

The problem should be practically relevant!

PreunderstandingPractical

get your feet wetempirical work, e.g., interview study or observation (participation)

Theoreticalscan ”relevant” literaturetalk to other researchersget a big picture of existing knowledgeensure theoretical relevance

11



S200110U-VI11(21)

3. Innovate

Hevner et al. 2004

12



S200110U-VI12(21)

4. Test / Validate

Innovate and test phases can be and often are intertwined

Validation is perhaps the hardest part of constructive researchValidation should be performed in industrial settings, whenever possible – to ensure practical relevanceValidation employs other techniques, such as action research and case studies

13



S200110U-VI13(21)

Market-based Validation of Managerial Constructions

Weak market testa manager applies the construction in a company

Semi-strong market testconstructions becomes widely adopted by companies

Strong market testsystematic application produces better financial results

Semi-strong and strong market tests require statistical analysis of a substantial amount of implementation data

14



S200110U-VI14(21)

Shaw/Validation

Shaw, 2001

15



S200110U-VI15(21)

Hevner et al. 2004

16



S200110U-VI16(21)

5. Show theoretical contribution & novelty6. Examine scope of applicability & generalize

Noveltycrucial, but don’t be too hard on yourselfLots of possibilities

entirely new idea (rare)cross-domain knowledge sharingimproved idea / implementation / solutioninteresting research approach...

Knowing the field & positioning is crucial to novelty and theoretical contribution

Generalizebroad = good?hypothesize as ground for further testing

17



S200110U-VI17(21)

Evaluation Criteria for Constructive Research

ConstructRelevance

Theoretical RelevancePractical Relevance

NoveltyPractical utility

difficult to assess the practical adequacy of any new construction prior to its implementationdifficult because of organizational factorstechnical success != practical success

Research processrigor

18



S200110U-VI18(21)

Hevner et al. 2004

19



S200110U-VI19(21)

Shaw, 2001

20



S200110U-VI20(21)

Shaw, 2001

21



S200110U-VI21(21)

Shaw, 2001

22



S200110U-VI22(21)

Shaw, 2001

23



S200110U-VI23(21)

Shaw, 2001

24



S200110U-VI24(21)

Shaw, 2001

25



S200110U-VI25(21)

Is the Constructive Approach Scientific?

Characteristic features of the constructive methodstep-by-step procedure, where steps can be checkedserves some definite purpose, is goal-driven

Objectivity, criticalness, autonomychecking the steps

Progressiveness, criticalnessshows concretely, which solutions work, and don’t workworking constructions tend to lead to new questions

Relevant, simple, easy to useinadequate solutions become eliminated by usersoften the simplest idea is the most adequate one

26



S200110U-VI26(21)

The SoberMethod(?)

The common methodological framework for SoberIT?Are we the ones embarking on the ”amalgamation”movement (Glass et al., 2004) at HUT?

”Although historically [computing] has evolved as several stovepipes of knowledge—predominantly, as we have said, CS, SE, and IS—there is now some impetus for amalgamation. ... some integrated schools of computing have already been formed, ...””There are problems on the amalgamation horizon, ... They have not, in the past, communicated well with each other. ... In what may be the biggest problem of all, there is a tendency for each of the fields to disdain the work of the others. ... These problems must be addressed before any amalgamation could possibly be effective. ”

27



S200110U-VI27(21)

Selected ReadingGlass RL. The Software Research Crisis. IEEE Software 1994;11(6):42-7.Alan R. Hevner, Salvatore T. March, Jinsoo Park, and Sudha Ram.

2004. Design science in information systems research. MIS Quarterly Vol. 28 No. 1, pp. 75-105/March 2004.

Kasanen, Eero, Lukka Kari, and Arto Siitonen. 1993. The Constructive Approach in Management Accounting Research. Journal of Management Accounting Research, 5 (1), pp. 243-263.

Shaw, M. 2001. The Coming-of-Age of Software Architecture Research. Proceedings of ICSE-2001, pp. 657-664. Los Alamitos, CA: IEEE Computer Society Press.

Shaw, M. 2002. What makes good research in software engineering?International Journal on Software Tools for Technology Transfer. (2002) 4: 1–7

Robert L. Glass, V. Ramesh, and Iris Vessey. An analysis of research in computing disciplines. Communications of the ACM Volume 47, Number 6 (2004), Pages 89-94.

28



S200110U-VI28(21)

Exercise

Read at overview level the section ”Guidelines for Design Science in Information SystemsResearch” of Hevner et al., 2004Study at surface level (abstract, intro, headings, conclusions) the example articles. What elements of the research cycle do you identify, how are they carried out, how well?Discussion

Hevner et al. 2004

Documents

Software Business and Engineering Institute