Statistics for the Rest of Us: Mastering the Art of Understanding Data Without Math Skills (Advanced Thinking Skills Book 4)

clean. Cleaning data is exactly what it sounds like: looking through it to make sure it doesn't have any errors or duplicates, that it was collected accurately, and that it's in the right format for your computer (or you) to analyze.[xvi]

five-step process: Define the question and figure out methodology; collect data; clean and summarize the data using descriptive statistics; process the data and apply hypotheses; make inferences and apply findings.[xii]

1) Strength of association (how strong is the association between the data?) 2) Consistency (will the same results be found if different people replicate the study at different times?) 3) Specificity (how specific is the association?) 4) Temporality (did the effect occur after the cause?) 5) Biological gradient (is there a correlation between the a

Another way of thinking about it is that p-values tell us how likely it is that another experiment would have the same results.

Bayes' Theorem looks at how likely something is to happen, given that something else has already happened.

Pitfall #2: Looking at the wrong measure of center

P-values are a probability and thus are expressed as a number between zero and one. Lower p-values mean that results are more statistically significant; higher p-values mean the results may be due to chance.

Pitfall #5: Getting causation backwards

This chapter will look at five common pitfalls in statistics and how you, the average consumer, can recognize them.