An example of evaluating an MBA proposal research method design in terms of research methods design quality criteria

An example of evaluating an MBA proposal research method design in terms of research methods design quality criteria (information from perplexity.ai)

My question: "Please comment on the research method design quality criteria (e.g. validity, reliability, dependability, etc.) of my MBA research design:

To study the the impacts of corporate downsizing on the staff morale in the Hong Kong banking sector

To conduct narrative interviews on 5 middle managers that have experience corporate downsizing in the Hong Kong banking sector so as to learn the downsizing impacts on staff morale. These 5 middle managers are my banking firm colleagues. I will also conduct an online semi-structured interview with my friend who is a university lecturer on Organizational Behavior to cover my research theme. The interviews with my banking firm colleagues are to be conducted in a Starbuck coffee shop in Hong Kong; the duration is 45 min. for all the interviews, whether online or in the Starbuck coffee shop.

The reply: "The research design employs narrative interviews with a small sample of colleagues and an expert, which suits exploring lived experiences of downsizing impacts on morale but raises concerns in several quality criteria for qualitative inquiry.

​

Credibility

Credibility, akin to internal validity, assesses if findings authentically reflect participants' experiences. Narrative interviews enable rich, personal stories from middle managers who endured downsizing, enhancing credibility through direct relevance. However, the expert interview adds theoretical depth, though its semi-structured format and single instance limit triangulation without further participant checks or peer debriefing.​

Transferability

Transferability evaluates if findings apply beyond the sample, similar to external validity. The purposive sample of five colleagues from one Hong Kong banking firm, spanning ages 30s-60s, offers contextual thickness for that sector but lacks diversity in firms or roles, hindering broader applicability. Detailed descriptions of the Starbucks setting and 45-minute duration aid reader judgment, yet no demographic or firm variety weakens this criterion.​

Dependability

Dependability parallels reliability, ensuring consistent processes if replicated. Standardized 45-minute durations and semi-structured protocols promote auditability, but researcher-participant familiarity (colleagues) risks bias, and no mention of audio transcription, coding protocols, or audit trails undermines auditability. Reflexivity on the female researcher's 30s age and colleague ties is essential but absent here.​

Confirmability

Confirmability checks if findings stem from data, not researcher bias (like objectivity). Proximity to participants as colleagues introduces power dynamics and social desirability risks, especially in a public Starbucks venue. The online expert interview mitigates some bias via distance, but without verbatim transcripts, member checking, or negative case analysis, confirmability remains low."

A collection of blog notes on using chatgpt for research purpose.