MMCA 

In papers that statistically link sensor data with collaborative constructs, researchers often begin with a construct(s) of interest and a data source. On one end, constructs are operationalized into measurable outcomes of interest, e.g., number of questions solved, collaboration quality scored by researchers, or the perceived helpfulness of group members on a scale of 1-5.. On the other end, from high-frequency data, metrics are generated as a hypothesized indicator of the target outcome, e.g., joint visual attention, vocal pitch, or the prevalence of certain hand gestures. Outcomes are often measured through manual methods (e.g., self-report, human coding), while metrics are usually calculated computationally.

For example, data in MMCA research could be the x, y, z location of gaze locations captured by an eye tracker. One metric from this data could be the amount of time two people spent looking at the same on-screen location. Coordination, a collaborative construct, could be operationalized as a score given by researchers on coordination efficiency, thereby becoming an outcome. The research concludes that gaze patterns are connected to coordination by showing a significant association between the gaze metrics and researcher scores.