What are exactly the steps to aggregate individual-level data at the group-level?
Let’s imagine you asked employees of different companies to evaluate the leadership style of their CEO and their own job satisfaction. In the end, you have 200 employees from 50 companies(or CEOs). For some companies you got only one employees and from some other you have 6 employees.
Usually, the leadership literature suggest to aggregate the CEO leadership style scale (5points Likert scale with 10 items) at the group level due to the interdependence of the employees who belong to that CEO.
How can I check ICCs criteria (for instance, inter-rater agreement or reliability, or rwg)? How should I aggregate and average the data about leadership style?
1. some companies or CEOs have only 1 respondent
2. 10 items scale for leadership style
3. 200 respondents are the rows of the actual data set and the items of the scales are the columns.
Thank you in advance for your support!