Even though most voters say that the case against the former president is “strong,” they don’t want to see him serving jail time
Small sample size is fine when it’s representative of the population. Trying to extract nationwide sentiment, a very diverse thing, off a small size is unlikely to be very representative.
Except the numbers work out, and studies have made very very sweeping generalizations based on much smaller sample sizes of much larger demographics (for example the 1 in 5 myth comes from a study that had less than 100 respondents). This study is a dream compared to those.