We conducted a semester-long ethnographic study of group work in a physics class that implemented the investigative science learning environment approach. Students' conversations were videotaped while they were engaged in group learning activities. Our primary research goal was to better understand what factors made some groups more effective than others. We developed a coding scheme that uses particular modality markers such as hedging and upward inflection to identify how students "position" themselves in these group conversations. Additionally we quantified group effectiveness by how many key ideas in a particular activity a group negotiated and resolved through the course of their conversation. This research builds on theories of social positioning that posit that groups are more effective when their discussion is more equitable. Our exploratory study indicates that groups whose participants position themselves in a more equitable way, are more effective at completing challenging physics activities and resolving areas of confusion that arise.