Two methodologies of database design research

Dan Scott, 2018-10-18

GLIS 703

Sukula, S. K. (2005). Indigenous database development in Indian research and development library and information centres. Online Information Review, 29(2), 193–207.

Sukula: problem statement

the Indian user seldom finds indigenous information in international databases [and] global databases are often biased towards Western publishing (2005, p. 194)

Sukula: epistemology


Positivistic research is often silent because it conceives itself as ``scientific'': the only valid approach" (Hjørland, 2000, p. 526)

Sukula: research question

  • Derived from the conclusions:
    • What actions should Indian libraries and information centres take to improve their development and promotion of indigenous databases?

Sukula: aim

to nurture and develop potential [indigenous] databases, and facilitate increased user access(2005, p.194)

Sukula: objectives

  • To study the development and maintenance of indigenous databases.
  • To discover indigenous database marketing practices.
  • To assess database quality and impact on library and information services.
  • To discover the possibilities for database use in an Indian context.
(2005, p.194)

Sukula: methodology

  • Quantitative
  • Method: descriptive survey (hypothesis free)
  • Instruments: questionnaire and interview
  • Sampling: purposive - chief librarians

Sukula: variables

  • Operationalized from the objectives
  • Five-point Likert scales measure agreement
  • Dichotomous and ordinal variables for other data

Sukula: analysis

  • Purely quantitative, rather simplistic:
    • Data tables (n & %)
    • Text summarizing the numeric data in each table
  • Concludes with strong ungrounded recommendations

Sukula: ethics

  • Good: Participants cannot be associated with their responses
  • Problematic: "the use of marketing techniques will not place an extra burden on existing library staff" (2005, p. 205)

Sukula: critiques

  • Chief librarians: source of problem and solutions
  • Library worker voices
  • References: dates, sources
  • Intro + conclusion vs. data

Sukula: further positivist evidence

  • No discussion of researcher bias; assumes an objective reality
  • Generalizes results drawn from Northern India to all of India

Godbold, N. (2009). User-Centred Design vs. “Good” Data Base Design Principles: a Case Study, Creating Knowledge Repositories for Indigenous Australians. Australian Academic & Research Libraries, 40(2), 116–131.

Godbold: problem statement

[creating a database shell for] clans to independently record their own traditional knowledge in the context of their unique knowledge system(2009, p. 117)

Godbold: epistemology

Subjectivist / transactional

results of the investigation are a product of interaction between the subject and the investigator(Pickard, 2012, p. 7)

Godbold: research question

  • Not explicitly stated, but implied:
    • How can user-centred design and standard database design principles accommodate traditional knowledge?

Godbold: aim

Driven by applied research:

[reflect on the process of] creating a robust, flexible, searchable infrastructure to support the archiving and ongoing use of the traditional knowledge being gathered by indigenous community members(2009, p. 117)

Godbold: methodology

  • Qualitative
  • Method: Case study? No: action research
  • Instruments: Human instrument: reflection, dialogue, meetings, usability tests
  • Sampling: Purposive (determined by applied research)

On action research

[action research is] an interventionist approach to research taken with the explicit intention of improving practice, and understanding that practice and the situation in which it takes place(Pickard, 2012, p. 158)

Godbold: analysis

  • Systematic reflection--examples:
    • "At first glance […] the TKRP project appeared to be a good candidate for creating data repositories which can later be shared" (2009, p. 118)
    • Generic personas cannot satisfy users from “several hundred language groups with all the cultural variety that implies” (2009, p. 119)

Godbold: analysis (2)

  • Fundamental contradiction:
    • A metadata scheme "designed to be a private, local solution for individual communities" cannot also support a merged public database (2009, p. 119)
    • A feature, not a bug!

Godbold: ethics

  • Was informed consent obtained from the participants?

Godbold: critiques

  • Mixed reporting: technical features + ethnographic insight
MethodsDescriptive surveyAction research
InstrumentsQuestionnaire, interviewsReflection, dialogue, meetings, usability test
EthicsObjectively fineHopefully fine