Once the coding scheme was determined, all transcripts
were independently reviewed a second time and coded in
their entirety according to the newly developed coding
scheme. To ensure coding reliability, the research team met
to discuss coding interpretations. Coding discrepancies were
discussed until consensus about the appropriate code was
obtained. Another team member randomly selected 10% of
the transcripts and recoded them independently again in terms
of the existing descriptive codes. In comparing the coding of
participants’ responses between this team member and the
other investigators, the Cohen’s kappa coefficient was 0.94.
Validity was addressed by checking transcripts against themes
and interpretations.