Polling data reveal that an overwhelming majority of nine-year-olds can correctly identify the logos of major cigarette brands.
Conclusion
There is little or no connection between recognition of cigarette brand logos and smoking.
Evidence
An overwhelming majority of 9 year old's can identify the logos of major cigarette brands. However, less than 1% of those 9 year old's smoke.
Evaluate
How could we argue that, despite this evidence, there is a connection between recognition of cigarette logos and smoking?
We could say,
We could also say,
Goal
Our objections were of two different forms: 1. just because there isn't a visible connection yet, doesn't mean there won't be one eventually (i.e. the author is judging something using too narrow a snapshot of time)
2. just because there isn't a visible connection for 9 year old's doesn't mean there isn't one for other age groups. (i.e. the author is judging something using too narrow a sample of age group)
One way or the other, they're both Sampling flaws. So we should look for an argument that is judging something broad in the conclusion, but relying on a sample that only represents a narrow slice of people / narrow slice of time.