You may test the correlation between perceived management style and manager-owner relationship in small business units. You can prepare a questionnaire for each variable, distribute it, analyse the respondents`s answers, and test the correlation.
You are creating a quantitative study but the language in your research question seems to be qualitative. Further, the word "perceived" seems traditional of a qualitative phenomenological study. However, if you want to capture the perception of a groups of individual, think about changing your research from quantitative to qualitative. If so, then you can create an interview protocol strategy that will allow you to create an alignment between your research question, you methodology, and your goal.
Creating your own interview protocol will allow you to tailor the questions, in a way that you will be able to target specific spectrums in your answer, allowing the individuals to respond, based on their perception, but based on their experience. You can also take that information and convert it into percentages, which will also allow you to continue using a quantitative method.
I will recommend you to make a case study (or case studies) about real firms and as Iberakis Faltas said - to create your own interview with open questions.
Rajapriya is correct, I have chaired at least a dozen dissertations that used the MLQ to respond to questions about perceptions of employees as to the leadership style of their managers.