# How do you evaluate statistical claims for alternative explanations?

###### Author

Evaluating statistical results or claims for alternative explanations can be tricky, but with some key tips, you can become more proficient at making sense of the numbers you encounter in your daily life.

1) Understand the context: Before diving into the numbers, it’s crucial to understand the context of the study or claim. What is the main objective? Who conducted the study? What are the potential biases or conflicts of interest? Knowing the context can help you be more critical of the findings.

For example, if a news report says that a new drug reduces heart attack risk by 50%, knowing who funded the study is essential. If the drug manufacturer funded it, the research could have biases.

2) Check the sample size: Sample size is the number of observations (people, animals, or objects) in a study. The larger the sample size, the more reliable the results are likely to be. Small sample sizes can lead to misleading or incorrect conclusions.

For instance, if a blog post claims that 80% of people prefer a particular brand of toothpaste, but only 10 people participated in the survey, the results may not be representative of the general population.

3) Assess the sampling method: Sampling methods can significantly impact the validity of a study. The ideal sampling method is random, meaning that each individual has an equal chance of being included in the study. Non-random sampling methods may introduce bias and reduce the study’s generalizability.

For example, a social media post that claims a specific diet leads to weight loss might only include people who already follow that diet, which would not provide a representative sample of the general population.

4) Examine the effect size: Effect size is a measure of the magnitude of the difference between two groups. A larger effect size indicates a more substantial difference. When evaluating a claim, consider the effect size alongside the statistical significance to better understand the practical implications.

For instance, a study might show that a new sleep aid helps people fall asleep 5 minutes faster than a placebo. While this might be statistically significant, the practical significance (effect size) is relatively small.

5) Look for confounding factors: Confounding factors are variables that can influence the relationship between the variables of interest. These factors can lead to incorrect conclusions if not accounted for in the analysis.

For example, a study might find that people who drink more coffee have a higher income. However, factors like education and occupation, which could also influence income, should be considered before drawing conclusions about the relationship between coffee consumption and income.

6) Assess the measure of central tendency: When looking at a dataset’s average (mean), it’s essential to consider other measures like the median and mode. These measures can give you a better understanding of the data’s distribution and help you assess the potential impact of outliers.

For example, a product rating might have an average of 4 stars, but if most ratings are either 1 star or 5 stars, the median and mode would provide a more accurate picture of the overall satisfaction of customers.

By keeping these tips in mind, you can become more adept at evaluating statistical claims and understanding the underlying data. This will help you make more informed decisions based on the information you encounter in news reports, blog posts, social media, advertisements, and product ratings.