“…Because of the time and expense involved in the coding group meetings, only 25 percent of each meeting was coded. The strategy to employ a subset of a group's communication exchanges has been validated [6] as well as employed in past CMCS research (e.g., [90,101]). Billings et al [6] reported validity coefficients of 0.8 or better for four-category interaction process analysis (IPA) coding in a 20 percent subset of the total interaction.…”