Foundations of Statistical Inference Flashcards
Probability
The likelihood that any one event will occur, given all the possible outcomes
Implies uncertainty - what is likely to happen
Essential to understand inferential statistics
Sampling Error
The sample mean won’t equal the population mean
Measured by the standard error of the mean
- If you repeat the study using new samples from the SAME population, how much will the sample mean vary?
Standard Error of the Mean
Basis for statistical inference
Allows us to estimate population parameters
Point Estimate
A single value that represents the best estimate of the population value
Many times it is the mean
Confidence Interval
A range of values that we are confident contains the population parameter
Width concerns the precision of the estimate
Correct Interpretation of Confidence Intervals
95% CI = if we were to repeat sampling many times, 85% of the time out confidence interval would contain the true population mean
Incorrect Interpretation of Confidence Intervals
There is a 95% probability that the population mean falls within an obtained confidence interval
The population mean is a fixed unknown value
Hypothesis Testing
Estimation of population parameters is only one part of statistical inference. Also used to make inferences about observed difference or apparent relationships from sample data
Alpha
Maximum probability of type 1 error
Set by researcher before running statistics
Usually set to 0.05 (max chance of type 1 error = 5%)
P-Value
Probability of type 1 error if the null hypothesis is true
Probability of observing a value more extreme than an actual value observed, if the null hypothesis is true
Type 1 Error and Significance
Mistakenly finding a difference (false positive)
Interpreting probability values - the p value is the probability of finding an effect as big as the one observed when the null hypothesis is true
Type 2 Error and Power
Mistakenly finding no difference (False-negative)
Probability of making a type 2 error
Statistical power - power is the probability that a test will lead to rejection of the null hypothesis, or the probability of attaining statistical significance
Two-tailed Test
Allows for possibility that difference may be positive or negative
One-tailed Test
More power = more likely to find significance when there is significance
Should only be used when the relevant difference is only in one direction
Statistical Power
The probability of finding a statistically significant difference exists in the real world
The probability that the test correctly rejects the null hypothesis
Only matters when the null is false