Unit 3 Lesson 1 - Bias and surveys Flashcards
open q’s
Open questions - usually in a text box. People can say anything they want really. Examples include “tell us about your experience at this restaurant”. Questions are called dirty because they can give answers one may not have thought of when creating the survey. Answers can be very creative.
closed q’s + 4 types of
Closed questions - never in a text box. Answers are clicked out of options.
Information questions - “how old are you?”
Checklist questions - you can answer as more than one answer. Ex “what did you have for breakfast?” you could select options such as cereal, fruit, bread, etc.
Ranking questions - you rank certain items from the same category. Ex. rank your favorite school subjects on a scale from x to x.
Rating questions - you rate something also on a scale. For example, rate your experience at this restaurant on a scale from 1 to 10.
what is bias?
Bias - an unintended influence on a data gathering method.
what is sampling bias
Sampling bias - selecting the wrong group of people to survey to represent an entire group of people. Ex. surveying only the girls in a class about a general question that affects everyone. This is unfair because everyone should be involved equally in surveys for the most part. If the survey were about female hygiene products then it would be suitable for only girls to participate in but if it’s a question about school or something then it is unfair since not everyone is fairly represented.
what is non-response bias
Non-response bias - a bias in which a sample is collected in such a way that some members of the intended population have a lower or higher sampling probability than others.
Ex. asking small children and older people to answer a survey on Twitter. This is unfair because these demographics are not as handy with technology as other generations and are not as likely to participate in the survey which leads to less people filling out surveys.
what is household bias
Household bias - When groups of different sizes are polled equally leading to under/overrepresentation.
When one person is asked a question and represents many or some others as well. A smaller household wouldn’t affect the opinions of one person but a big household with many other opinions may
what is response bias
Response bias - the tendency of a person to answer questions on a survey untruthfully/misleadingly.
Ex. if when someone is taking a political survey, they feel pressured to choose a certain option to be considered conventional by society.
Mistakes in data representation - improper scales
Improper scales - can lead to cherry picking results to prove points.
ex. homicide in Ontario but the graph made it look fine
Mistakes in data representation - % in small sample space
Percents in small samples - choosing a small sample size and then using a percentage to make the numbers look good.
Ex. asking 2 students if they like the new cafeteria food. If both say yes, then 2/2 as a percentage is 100%. Then the surveyor could say that 100% students enjoy the food at the school when in reality only 2 people were surveyed and those 2 people don’t represent the entire population.
Mistakes in data representation - misleading titles
Misleading titles - titles that don’t make sense or that give the reader a certain impression about data. great link here
