Technology-supported citizen science has created huge volumes of data with increasing potential to facilitate scientific progress, however, verifying data quality is still a substantial hurdle due to the limitations of existing data quality mechanisms. In this study, we adopted a mixed methods approach to investigate community-based data validation practices and the characteristics of records of wildlife species observations that affected the outcomes of collaborative data quality management in an online community where people record what they see in the nature. The findings describe the processes that both relied upon and added to information provenance through information stewardship behaviors, which led to improved reliability and informativity. The likelihood of community-based validation interactions were predicted by several factors, including the types of organisms observed and whether the data were submitted from a mobile device. We conclude with implications for technology design, citizen science practices, and research.