1) Always based on haphazard sampling. There is no attempt to sample in a statistically sound manner which could include all possibilities. Anecdotal evidence illustrates only one possibility. So, there is no way to evaluate how frequently it occurs compared with other possibilities. Anecdotal evidence is often gathered from self selected sources. Anecdotal evidence should not be confused with statistical studies which do not fully document their sampling methods.
2) Always has a small sample size usually an individual case. This is the reason why it illustrates only one possibility to the exclusion of others. This should not be confused with a study which focuses on a narrowly defined group. For example, a well designed study to determine the presidential candidates of women in Gallop New Mexico would not be anecdotal even though it is limited to a specific group..
3) Always designed to have emotional impact. Anecdotal evidence is used because it has great power to convince.
4) Usually from a source with a hidden agenda or bias. Anecdotal evidence has a definite purpose. It is always offered in support of a cause, hypothesis or agenda which usually is not directly stated.
Summary: While an anecdote may be true, it often is not representative. It's deliberately designed to be an overly influential outlier.
4. Anecdotal evidence is often used by politicians and popular magazines to support their point of view.