Quiz 4 Flashcards
normal distribution
Distribution of sample means
Standard normal distribution
Specific type of normal distribution with a mean of 0 and a standard deviation of 1
68-95-99% Rule
Approximately 68% of the data falls within one standard deviation of the mean, 95 within 2, and 99 within 3 in both directions
Point of 65-95-99 rule
Constructing confidence intervals
Confidence intervals
Range of values into which a population parameter is likely to fall for a given level of confidence.
For a 95% CI, add and subtract 1.96 (2 standard deviations of the mean)
ONLY express sampling error
What do we get from testing the null hypothesis
A standard of evidence that must be met to reject it, which is expressed as alpha set to 0.5 (significance level)
Significance level
Chance the evidence we observe would occur randomly if the null were true
When is something statistically significant?
If it’s unlikely that the null is true
Null Hypothesis Significance Testing
1: propose a research hypothesis
2: set the significance level (susally 0.05 percent)
3:Estimate parameters using sample data
4:Calculate CI or p-value
5: Reach a conclusion about null
P-value
The smallest standard of evidence at which we would reject the null. At a 5% significance, we reject the null if p < 0.05
Significance Testing with Confidence Intervals (steps)
1: Identify hypotheses
2: Set confidence level
3: Calculate CI
4: Is null within interval
5: Conclude
CI formula
Sample mean +/- z * standard error
Case Study Design
Involve either a single case or a small number of cases. Uses comparison to make inferences about relationships
Method of difference
Reseachers select cases where the outcomes differ, compare cases, and conclude causal factors
Method of agreement
Selecting cases with the same outcome and identity the things they have in common