Self-reported studies are not very useful in collecting usable data, they are typically utilized to confirm biases.

This kind of online polling has no way to control variables that can invalidate the premise. Were these questions translated, or are we asking 22k that could afford a secondary education if they are happy? What is the demographic of people who have access to the Internet in the country? Did we just ask 22k people attached to the royal family of Saudi Arabia if they were happy?

This "study" raises more questions than it attempts to answer.