T test of means
Key Assumptions in Hypothesis Testing
Alternative Hypothesis ((H_a))
Defines whether (\mu) (population mean) is either less than, greater than, or not equal to (\mu_0) (hypothesized mean).
Assumption 1: Simple Random Sample (SRS)
Must ensure sample is randomly selected.
Assumption 2: Population Size
Population must be greater than 10 times the sample size (n).
Assumption 3: Sample Size Conditions
Large Sample Size: If ( n \geq 45 )
This supports the Central Limit Theorem.
Distribution's skewness or outliers do not impact the results.
Medium Sample Size: If ( 15 \leq n \leq 45 )
For moderate skewness, a maximum of one to two outliers is permissible.
Data can be symmetric or moderately skewed.
Small Sample Size: If ( n \leq 15 )
Data must be symmetric with no outliers.
Evaluating Skewness in Sample Data
To demonstrate medium or small sample conditions:
Visual Tools:
Box-and-Whisker Plot:
Effective for showing outliers.
Can indicate skewness visually.
Histograms:
Useful but not as effective for outlier detection.
Judgment in Samples:
If sample size is 44 or equal to 45, judgment call on skewness is needed.
A larger sample size can accommodate more skewness.
Lower sample sizes require stricter adherence to symmetry guidelines.
Choosing Between One-Sample T-Test and Z-Test
Z-Test: Used when the population standard deviation is known and sample size is large (( n \geq 45 )).
T-Test: Used in cases where population standard deviation is unknown:
The formula utilized:
( t = \frac{\bar{x} - \mu}{s / \sqrt{n}} )
Involves sample standard deviation (s) instead of the population standard deviation (( \sigma )).
Degrees of Freedom:
Calculated as ( n - 1 ): represents the number of independent pieces of information.
Practical Application of the T-Test
Probability Determination:
Finding (t) scores using tables or statistical software to determine critical values or p-values.
Example Scenario:
If given a ( t ) score of 1 with 10 degrees of freedom, understand how to interpolate between values in a t-table to determine probabilities.
Conclusion and Exploration of New Tools
Familiarity with statistical software or calculators can ease calculations.
Discussion of visual aids and their importance in assessing data distributions.
Suggested exploration of additional learning resources or videos for clarity on concepts discussed.