Explore: Statistical Questions and Bias Solutions

Statistical Questions and Bias Solutions

A statistical question has an answer that allows for variability . 
In the study of statistics, variation occurs naturally and normally in data collections.  
However, the results of data collection must be repeatable to show there is no bias (or little bias).  
If there is little to no variability, it is possible that the sample: 
- Does not accurately reflect the population
- Includes bias and/or errors  
Statistical bias is an error that causes the results of a study to misrepresent any population parameter .  
When bias is present, the collected data is not representative of the population (or sometimes even the sample).  
Bias can occur when questions:  
- Are leading
- Are inaccurate
- Do not allow for variability
Therefore, statistical questions should not: 
- Include the opinion of the writer/researcher
- Lead individuals of the population to answer in a particular way
A purpose of study is a simple statement that is clear and without presumption that explains the goal of the research. 
- When writing statistical questions, a purpose of study can help you write questions that ensure you collect the right data without bias . 
- When participating in a study, identifying the purpose of study can help you be a healthy skeptic and recognize bias .

Describe the bias in each question.

Sample: The result is one single number, which means there is nothing to analyze. 

Everyone knows that schools with mascots are better than schools without. What are your reasons for agreeing?

Sample: You are only allowed to give an answer that agrees with the statement. “Everyone knows” implies consensus.

Rewrite the question to remove bias. 

Given the obvious benefits of raising the age at which you can get your license, what are your thoughts? 

Sample: What are your thoughts about raising the age at which you can get your license? 

Sample: What are the pros and cons of raising the age at which you can get your license?