Robert Pirsig wrote a whole book on the nature of Quality...
In the evolutionary sense, "better" means "more fit for survival". Which is to say it's an empirical test. The adaptations which are the "best" are those which survive to create offspring.
In human affairs the question becomes more complex, and can be subjective. In part it depends on the environment and success criteria you've established. Are you trying to inform? To entertain? To be commercially successful? To pitch your startup? To expand your own understanding?
Not quite writing, but related: Jacob Nielsen's approach to Web UI testing is one that I've long been impressed by. Rather than dissect a design into its components and assess those individually, he has a very simple, and effective, test: assess the suitability to task of the interface. Define what it is you want the user to do and let them attempt to do it, using different designs. See which is the most successful. After you've identified which is better, you can look at the differences between the designs (or texts) and figure out why one succeeds where the other doesn't. It's simple, highly empirical, and, by focusing on goals rather than means wraps up a huge bundle of variables that would otherwise cloud assessment.
Robert Pirsig wrote a whole book on the nature of Quality...
In the evolutionary sense, "better" means "more fit for survival". Which is to say it's an empirical test. The adaptations which are the "best" are those which survive to create offspring.
In human affairs the question becomes more complex, and can be subjective. In part it depends on the environment and success criteria you've established. Are you trying to inform? To entertain? To be commercially successful? To pitch your startup? To expand your own understanding?
Not quite writing, but related: Jacob Nielsen's approach to Web UI testing is one that I've long been impressed by. Rather than dissect a design into its components and assess those individually, he has a very simple, and effective, test: assess the suitability to task of the interface. Define what it is you want the user to do and let them attempt to do it, using different designs. See which is the most successful. After you've identified which is better, you can look at the differences between the designs (or texts) and figure out why one succeeds where the other doesn't. It's simple, highly empirical, and, by focusing on goals rather than means wraps up a huge bundle of variables that would otherwise cloud assessment.