Qualitative researchers are required to articulate evidence of four primary criteria to ensure the trustworthiness of the study’s findings: credibility, transferability, dependability, and confirmability.
Credibility corresponds to the notion of validity in quantitative work but is more about internal validity. The credibility of qualitative data can be assured through multiple perspectives throughout data collection to ensure data are appropriate. This may be done through data, investigator, or theoretical triangulation; participant validation or member checks; or the rigorous techniques used to gather the data.
Transferability is like generalizability in quantitative; however, it is not generalizability. Transferability addresses the applicability of the findings to similar contexts or individuals not to broader contexts. Transferability can be achieved by a “thick description” of the findings from multiple data collection methods.
Dependability is like reliability in quantitative studies. Dependability can be ensured through rigorous data collection techniques and procedures and analysis that are well documented. Typically, an inquiry audit using an outside reviewer assures dependability. For students, this would be your committee.
Confirmability is like objectivity in quantitative studies; however, objectivity is not necessarily critical for qualitative studies as long as personal biases are unpacked in the write-up. Unpacking personal bias can be accomplished by a bracketing interview or reflexivity. Confirmability of qualitative data is assured when data are checked and rechecked throughout data collection and analysis to ensure findings would likely be repeatable by others. Confirmability can be documented by a clear coding schema that identifies the codes and patterns identified in analyses. This technique is called an audit trail. It can also be ensured through triangulation and member checking of the data as well as conducting a bracketing interview or practicing reflexivity to confront potential personal bias.