A transparent research laboratory with an interview room on one side and an archival document wall on the other, a silver robot sorting coded cards between them

Chapter Three · Methodology

How the evidence was built

A transparent account of design, data, and limits — as submitted for the Doctor of Philosophy in Information Technology.

AI-generated conceptual illustration · Not research evidence or a real study site

Four glass domes on a steel table, each protecting a miniature workplace world
Seven organizational contexts — supply chain, banking, higher education, software, logistics technology, public service, and frontline manufacturing — imagined as one field of study.AI-generated illustration · Not research evidence or a real study site
Silver robot placing coloured glass coding tokens along ribbons of transcript light
Deductive and inductive coding stayed inside the established six-theme, thirty-code framework.AI-generated illustration · Not research evidence or a real study site

  1. 01 · DESIGN

    Exploratory qualitative study guided by Sensemaking Theory (Weick, 1995) — how individuals interpret information and construct meaning within organizational environments.

  2. 02 · INSTRUMENT

    Semi-structured interview protocol, expert-reviewed by doctoral faculty and field-tested before use.

  3. 03 · SAMPLE

    Twelve participants (P1–P12) across seven organizational contexts: supply chain and logistics, banking IT, higher education administration, software engineering, logistics technology, government human services, and frontline manufacturing. Recruited purposively through professional networks and LinkedIn, with five targeted follow-up recruits to reach underrepresented frontline and supply-chain roles.

  4. 04 · INTERVIEWS

    Conducted virtually on a secure platform, approximately 45 to 60 minutes each, audio-recorded with permission and transcribed verbatim.

  5. 05 · SECOND SOURCE

    A purposively selected corpus of publicly available NCDOT organizational documents — Annual Report, Performance Dashboard, and Employee Engagement Surveys — coded to the same six-theme framework, with dated copies personally retrieved and archived.

  6. 06 · ANALYSIS

    Deductive and inductive coding in Delve within the established six-theme, thirty-code framework; In Vivo coding preserved participants’ own language. No new themes or codes were created during analysis.

  7. 07 · EVIDENTIARY STANDARD

    Every quotation reproduced verbatim; thin, absent, or ambiguous evidence stated explicitly rather than inferred; causal language avoided in favor of qualitative hedges; divergent cases preserved rather than resolved into a single pattern.

  8. 08 · TRUSTWORTHINESS

    Credibility, dependability, transferability, and confirmability addressed through cross-participant comparison, an audit trail of coding decisions, and reflexivity and bracketing by the researcher.

Recruitment · Interview sign-up

Take part in an interview

Professionals who work with data at work — in supply chain and logistics, manufacturing, banking IT, higher education, software, or public-sector service — are invited to join a 45 to 60 minute virtual interview. Sessions are audio-recorded with permission, transcribed verbatim, and reported with fictional identifiers so no participant is identifiable. Signing up registers interest only; the interview itself begins with a review of consent, and you may decline any question or withdraw.

Sign up for an interview

Opens the official interview sign-up form in a new tab.

An illuminated map of findings ending at an uncharted edge of silver mist
Examining boundaries, alternative explanations, and what the evidence cannot settle.AI-generated illustration · Not research evidence or a real study site

Limitations, stated honestly

  • The qualitative design focused on rich descriptions of participant experiences rather than statistically generalizable findings; findings may not generalize to all industries, organizations, or employee populations.
  • Purposive sampling recruited participants with specific eligibility criteria rather than random selection, so the sample may not represent all employees in supply chain and operational environments.
  • The study relied on self-reported experiences collected through semi-structured interviews.
  • Full data saturation was not established, particularly within the frontline, shop-floor subgroup — two frontline interviews are not treated as establishing saturation for that subgroup.
  • Field notes were not produced as a separate, reviewable data source alongside the interview transcripts, so comparison against field notes could not be conducted.
  • Visible enthusiasm for the data storytelling concept expressed during several interviews is acknowledged as a possible source of participant agreement bias.