Statistical Inference Using T-distribution

Discussion begins with the need to develop alternative hypotheses after establishing initial hypotheses.
- Alternative Hypothesis (H1): Represents the hypothesis that the test aims to support or prove.
- Null Hypothesis (H0): The default position that indicates no effect or no difference, and must be maintained until evidence suggests otherwise.
Population and Sample Estimation
- Importance of estimating population parameters using sample statistics in hypothesis testing.
- Transition from population parameters to sample statistics.
- Population mean becomes a sample mean, leading to a shift in distribution representation.
- T-distribution is employed when the population variance is unknown and must be estimated from the sample.
- Characteristics of the t-distribution:
  - Symmetric and bell-shaped, similar to normal distribution but with heavier tails.
- Relates to degrees of freedom, which are critical when determining t-values for inference intervals.

The necessity for statistical inferences to be based on t-distribution when population variance is unknown.
Degrees of Freedom (df): Critical for interpreting t-values accurately.
- The computation of degrees of freedom needs to take into consideration the sample size, typically computed as:
  $ext{df} = n - 1$
  where n is the sample size.
Probabilities associated with the t-distribution are given for values that fall to the left or right of a specified t-value.
T-Table Utilization:
- Understanding how to reference the t-table.
- Locate the appropriate degrees of freedom for a given test statistic.

Calculating confidence intervals with t-values.
- Example given for finding t-values corresponding to degrees of freedom of 11.
- Critical Value: For example, locating 2.201 based on the t-table where df = 11.
When checking values for statistical significance:
- Compare test statistic with those found in the t-table to establish thresholds for hypothesis testing based on a defined alpha level.
- Example discussing critical value locations and related probabilities associated with t-distribution.

Discussion regarding significance levels and determining p-values from the t-test computations.
- If a test statistic (e.g., 2.24) falls below the critical t-value corresponding to a specific significance level (e.g., 2.351 for 1% significance), a hypothesis cannot be rejected.
Interpretation of p-values:
- p-value: The probability that the observed data (or more extreme) would occur under the null hypothesis.
- In this case, the critical threshold was referenced at 5% and 1%, leading to acceptance or rejection of null hypotheses based on strength of evidence.
Importance of assessing both critical value and p-values when conducting t-tests.
- For instance, if the observed p-value is computed as 1.34, the null hypothesis may not be rejected depending on pre-established thresholds.

Clarification on types of hypothesis tests:
- A one-tailed test focuses on deviations in one direction (greater than or less than).
- A two-tailed test considers deviations in both directions (not equal to).
- Rejection criteria differ based on the type of test due to the division of alpha throughout the critical regions.
Example mentions examining critical values at both extremes for acceptance or rejection of null hypothesis in a two-tailed scenario.
- Specific values (like 2.75) are cited as thresholds for making inference decisions.
- Rejection regions are established, indicating where null hypothesis should be rejected based on statistic outcomes (e.g., anything above 2.75 or below -2.75 is a rejection).
- Further discussion leads into numerical value examples with implications for hypothesis testing processes.