Define your terms specifically.
Bad: How does music affect people's focus? [What kind of music? How do you measure "focus"?]
Better: How does listening to classical music affect the length of high school students' study sessions? [You clearly defined the music, the people, and the focus, good!]
Bad: How are Russians reacting to their media's portrayals of the war in Ukraine? [Do you speak/understand Russian?]
Better: How are Russian expatriates in the west analyzing the Russian media portrayal of the war in Ukraine? [This is more likely to be in English, and accessible to you.]
Bad: What were the causes of the U.S. Civil War? [Many historians have written entire books on this subject.]
Better: How did the border conflict in Missouri and Kansas in the 1850s contribute to the eventual outbreak of the U.S. Civil War? [This is more focused in geography and time.]
If your question has words like good, bad, worst, best, etc., it is making a value judgment. These kind of questions are very hard to measure and thus very hard to answer in a research paper. If you want to evaluate something, make it something that's measurable.
Bad: Is it better for the U.S. to switch to a single-payer healthcare insurance system, like England's? [It's very hard to make one absolute statement on this because "better" can be measured by many different outcomes.]
Better: Which type of healthcare insurance system results in shorter hospital stays and better patient outcomes after major surgeries? [This is still an evaluative question, but you've quantified how you will measure "better".]
No questions with Google-able answers!
Bad: Has there been an increase in homelessness in the Dallas in the past ten years? [This is just a yes/no question]
Better: How have economic and political factors affected patterns of homelessness in the Dallas over the past ten years?
You are adding your perspective to an issue that is relevant to both your field of study and society at large. It should also be an issue that isn't answered by simple questions but has room for interesting debate.
Research is about informing, not dictating what X should do about Y.
Bad: What should the government do about low voter turnout? [It's unlikely that one single solution will increase turnout for all voters.]
Better: What are effective communication strategies to help increase voter turnout among under-30s? [You may still make recommendations, but it would be based on specific actions that have been shown to affect a specific group.]
Bad: How can doctors encourage higher Covid-19 vaccination rates? [It's unlikely that one or two actions will encourage vaccination for all people.]
Better: What factors most affect willingness to get a Covid-19 vaccination among the white, rural population? [Investigating and informing about one specific population.]
You don't have to ask something totally groundbreaking and new, but it should have some unique angle. If you are really interested in a study that someone else already did, see if you can find a unique angle by, for example, focusing on a specific location or population group of interest.