First, he/she must find the suitable statistical 'tools' for handling the specific problem, from a statistical point of view.

Then, to use these 'tools' without hesitating to ask for help, if something seems problematic.

How can a researcher do a good statistical analysis for a problem?

Generally the researcher should understand the objective of each statistical analysis & how to interpret the results generated by the analytical tools.  You can also refer to the following RG link.

https://www.researchgate.net/post/Which_statistical_analysis_do_I_use_for_data_analysis_of_a_questionnaire

First, he/she must find the suitable statistical 'tools' for handling the specific problem, from a statistical point of view.

Then, to use these 'tools' without hesitating to ask for help, if something seems problematic.

If unsure of how to proceed, the researcher should consult a professional statistician. He could explain what he tries to measure and what hypothesis he wants to prove to ensure that the experiment is well designed and that he will have sufficient data to guarantee the validity of the conclusion he will find...