Statistics

• Descriptive statistics

refer to the direct numerical measurement of characteristics of a population such as how many of something there are, what the average number of some phenomenon is, or what the range of a particular value of something is. I am describing what is there, but not going beyond the data. If I conduct descriptive statistics on all swans to test my hypothesis that all swans are white, I would have to describe every swan. Formally, a population is defined as a well-defined, complete collection of things, objects, and so on. A descriptive analysis requires a description of the entire swan population.

• Inferential statistics

comes to the rescue when I can’t measure all swans, because this approach allows me to measure a sample of swans, a subset of the swan population, and then make inferences or estimates about the population as a whole from the sample that was drawn. Inferential statistics solves the measurement dilemma as long, of course, as you follow some basic rules such as randomization and appropriate sample size.

Randomization allows researchers to make inferences about a population based on the way a sample is chosen. Every member of the population must have the same chance of being in the sample.

Another key ingredient to ensuring that your sample is representative of the population is sample size, the number or n of individuals in your sample

A sample size of one, from a statistical perspective isn’t likely to represent the population as a whole? Correspondingly, this is why most people are more likely to trust advice if the same data comes from multiple people.

• Thinking Statistically by Uri Bram
• How to Lie with Statistics by Darrell Huff
• Turning Numbers into Knowledge by Jonathan Koomey, PhD
• Principles of Statistics by M. G. Bulmer
• Statistics Final Exam - YouTube