Generally, this fallacy occurs when too few examples are used to prove a point about all members of a class or category.  The examples observed may be factual, but there are not enough of them to warrant a conclusion.  This should not be confused with a false extrapolation, which is a hasty generalization that involves a prediction.  The standard Dicto Simpliciter does not involve a prediction.

Two subcategories:

Fallacy of Accident:

Occurs when a generalization is made about a seemingly sensible rule, but this sensible rule doesn't work well in all situations.

Example: Stealing is bad. 

This is mostly true; however, what about stealing to feed your starving children?  Are there occasional exceptions to some generally agreed-upon rules?

Example: One should always return borrowed items when the person who loaned them wants them back.

Again, mostly true... unless that person, your neighbor, requests that you return the rifle he loaned you for hunting season seconds after he just threatened to kill his wife.

Example: Never talk to strangers.

Yup... unless, of course, you've fallen off of a cliff, broken your leg, and a stranger happens to be hiking past.

Faulty/Misleading Statistic:

Statistics can be misleading for a number of reasons.  They could be derived from an insufficient sample size (related to stereotyping), but they could also be taken from irrelevant sample populations.  Also, they could be misleading because of the way in which the numbers are presented.

Whenever statistics are presented, ask:

• Who is presenting those numbers and is there any reason this person might be trying to mislead?

• What is the source of the statistics being used?  Is that source reliable or biased?

• Are the statistics being presented relevant and do they prove what they are intending to prove?

• Do they make sense?  Are they understandable?

Example: "I once saw a TV show with an American who wore her shoes inside her house; therefore, all Americans wear their shoes inside their houses.

This is obviously an unfair stereotyping based on one example.  Even if the one example is true, it cannot reasonably be extended to represent 100% of the people who resemble that person.  This fallacy is often used to explain prejudice, and is usually perpetrated by people who have limited experience with those who they are prejudiced against. They use dicto simpliciters to fallaciously justify their prejudices because it feels logical.

Example: 100% of people who smoke die.  Smoking kills.  Don't smoke.

True.  But this claim implies that smoking causes all of these deaths.  According to the World Health Organization, "tobbacco kills up to half of its users." That's a lot but it's not 100%. This sort of gets into a causation vs correlation discussion, too.

Example:  According to The Snoop Newspaper's online poll, only 23% of the respondents think that QSI HPH is a good school.  Clearly, QSI HPH is a terrible school as 77% think it's an awful place.

While these numbers in this hypothetical example could be correct, the poll itself is not representative of all people.  Readers of this particular newspaper only constitute a small, specific portion of the population that might have a different position on the debate.  By and large, internet polls are NOT quality sources to use in a respectable debate. WHY?

Example: I'm shocked to learn that 50% of American students perform at or below the national average on this test!

While the number itself is obviously accurate, the statistic is framed as if it's shocking.  The framing is deceptive and leads to a faulty interpretation of the statistic.

Example: QSI HPH students are really lazy.  Did you know that 40% of their absences occur when those absences extend weekends by one day (on either a Monday or a Friday)?

Again, the framing of the number is at fault for the faulty interpretation of this statistic.  If students were absent randomly, they would have a 20% chance of being absent on any given day of the week.  Since being absent on both a Monday or a Friday would extend a weekend by one day, it stands to reason that, even in random circumstances, 40% of absences would extend weekends by one day.  Nobody should be shocked by this.

Example: Mr. Spagnolo threw out his back while reaching down to click the mouse on his computer; therefore, reaching to clicking a mouse on a computer is an extremely dangerous activity.

While the unfortunate back-throwing-out incident is true, it is only one incident and could be attributed to 1,000 other factors (namely the huge bench-pressing and squat session I completed five minutes before deciding to check my email).  This is an over-generalization and the inherent dangers of mouse-clicking cannot be determined from a sample size of only one incident. Additionally, this again gets into the causation vs correlation dicsussion.

Example: Don't you know that the life expectancy of people in the USA has risen from approximately 48 years old in 1900 to over 78 years old in 2010!  That means that people who used to only live to be 48 years old now live to be 78!

Actually, while we do live years longer than we did in 1900 because of medical advancements, a significant change to the statistic is the fact that the infant mortality rate is now much lower than it was in 1900.  Removing most of the "0"s and "1"s from the calculation of the average significantly raises the number.  Those who make it to adulthood in 2021 have more similar life expectancies to those who made it to adulthood in 1900 than the number suggests.

Example: An article titled: "Hidden Camera Proves the True Nature of Pit Bulls. Must See." was accompanied by this video:

Is that really any different than claiming that THIS VIDEO defines all cats or THIS MAN'S ACTIONS are reflective of the actions of all people from Wisconsin?

EXAMPLE: This TV Graphic

The numbers listed in this example were accurate; however, the visual representation of those numbers is not-to-scale and misrepresents a 6/7 ratio.