precision and confidence levels

Question 1

standard error

Answer 1

• Estimate of precision
• Smaller SE = more precise
• Makes the jump from sample to population
• A larger SD results in a larger standard error (more variability means more uncertainty)
• A larger n results in a smaller standard error (more people in the sample means less uncertainty)

Question 2

SD vs SE

Answer 2

• Standard deviation and standard error mean very different things
• SD relates to how spread out the values are in the sample collected (descriptive)
• SE relates to how precise our mean estimate is (inferential)

Question 3

how we use SE

Answer 3

• Use it to calculate many inferential statistics
• Significance tests
• Confidence intervals

Question 4

confidence intervals

Answer 4

• Estimate (mean) is not likely to be the ‘true’ estimate
• SE indicates how precise the estimate is
• Can use these together to give idea of what the true estimate is
• We attach a level of confidence to this idea
• Confidence intervals are linked to the central limit theorem
• If I was to take repeated random samples, 95% of my confidence intervals would contain the true population estimate
• Often use 95% however some fields use 99%
• All estimates of population values should be presented with a CI
• My sample estimate is my best guess, and I am 95% confident that the true population estimate is between these two limits

Question 5

how to calculate a confidence interval

Answer 5

Take mean and multiply by SE

Question 6

when can’t we use confidence intervals

Answer 6

• If the data is not normally distributed, we can’t use mean/SD
• If the data is ‘too’ heavily skewed, the normality assumption means we cannot use the SE

Question 7

SE assumptions

Answer 7

• Data is approximately normally distributed

* There is sufficient sample size (>20 individuals)