Flashcards on Understanding Statistical Fallacies and Misrepresentation
Understanding Statistical Fallacies and Misrepresentation
Tap to flip · Swipe to navigate
Statistical Misleading Techniques
15 cards
Card 1
Question: What is a 'sample with built-in bias' and why can it mislead statistics?
Answer: A sample with built-in bias is one that doesn't represent the whole population (e.g., only those likeliest to respond). It can make statistics mislead
Card 2
Question: Name two ways surveys can be biased without anyone directly lying.
Answer: 1) Certain groups are more likely to respond than others; 2) Respondents may give dishonest or inaccurate answers.
Card 3
Question: When evaluating a survey result, what three questions should you always ask about the sample?
Answer: Who was surveyed? How many people participated? Does the sample represent everyone?
Card 4
Question: What are the three types of averages, and how is each calculated or defined?
Answer: Mean: total divided by number of values. Median: the middle value. Mode: the value that appears most often.
Card 5
Question: Why can the mean be misleading when incomes are highly unequal?
Answer: A very high income for one person can raise the mean, making average income look high even though most people earn much less.
Card 6
Question: If someone reports an 'average' value, what should you ask to assess its accuracy?
Answer: Which kind of average was used (mean, median, or mode), and whether that average accurately represents the situation.
Card 7
Question: What problem arises when key details are left out of a statistical report?
Answer: Leaving out important information (like sample size or context) can exaggerate findings and make results unreliable.
Card 8
Question: Why is a small sample size a concern when judging a study's results?
Answer: Small samples may show results that appear significant but are unreliable because they lack sufficient data to support strong conclusions.
Card 9
Question: What three things should you check to determine if data are strong enough to support a conclusion?
Answer: Sample size, missing information, and whether the data are robust enough to justify the conclusion.
Card 10
Question: What does 'much ado about practically nothing' refer to in statistics?
Answer: Presenting tiny differences as important when they may be within the study's margin of error or due to random chance.