Hi everyone. I am looking for guidelines as to what is the cutoff % of missing data acceptable AT THE PARTICIPATION and at THE PARTICIPANT level for longitudinal data applied to a 2 level random slope random intercept MLM. Below are the details of my study:

Participants answered personality measures once a day up to 10 days. My research is interested at observing patterns of stability vs. change at the individual level (i.e., how much people vary over the course of 10 days in comparison TO THEMSELVES) and group level (i.e., how do X trait varies over 10 days for men vs. women for example). My problem with missing data is: 1- Some participants answered 75% of a questionnaire on a given day (for e.g.). What is the cutoff to indicate that it is sufficient to include (e.g., is only 5% of missing data allowed, 20%? On a given measure). 2- Some participants only answered 2, 3, 4 of the 10 questionnaires. What is the cutoff on how many missing data at the participant level is considered too much to include them? (e.g., if a participant only filled one questionnaire, their data is cross sectional, is that ok to include? what if they answered 3 questionnaires (i.e., 33%)?).

Any guidance will be appreciated!