Hypothesis Tests Regarding a Parameter: Structured Study Notes

Hypothesis Tests Regarding a Parameter

Introduction to Hypothesis Testing

Hypothesis testing is a fundamental statistical procedure used to make inferences about population parameters based on sample data. It involves making an assumption about a population characteristic and using evidence from data to assess the validity of that assumption.

• Hypothesis: A statement regarding a characteristic of one or more populations.

• Hypothesis Testing: A procedure, based on sample evidence and probability, used to test statements about population characteristics.

Steps in Hypothesis Testing

1. State the Hypotheses: Formulate the null and alternative hypotheses about the population parameter.

2. Collect Evidence: Gather sample data relevant to the hypothesis.

3. Analyze Data: Use statistical methods to assess the plausibility of the null hypothesis.

Formulating Hypotheses

Null and Alternative Hypotheses

Hypothesis tests always begin with two competing statements:

• Null Hypothesis ($H_0$): The statement to be tested, usually representing no change, no effect, or no difference. It is assumed true until evidence suggests otherwise.

• Alternative Hypothesis ($H_1$): The statement we seek evidence to support, representing a change, effect, or difference.

Types of Hypothesis Tests

Depending on the research question, hypotheses can be structured as follows:

• Two-Tailed Test: Tests for any difference (not equal).

• Left-Tailed Test: Tests for a decrease (less than).

• Right-Tailed Test: Tests for an increase (greater than).

	Two-Tailed	Left-Tailed	Right-Tailed
Null Hypothesis	$H_0: ext{parameter} = ext{value}$	$H_0: ext{parameter} = ext{value}$	$H_0: ext{parameter} = ext{value}$
Alternative Hypothesis	$H_1: ext{parameter} eq ext{value}$	$H_1: ext{parameter} < ext{value}$	$H_1: ext{parameter} > ext{value}$

Note: Left- and right-tailed tests are called one-tailed tests. The direction of the inequality in $H_1$ determines the tail.

Examples of Hypothesis Formulation

• Population Proportion: Claim: The proportion of children experiencing headaches with a new drug is different from 2%. Null Hypothesis: $H_0: p = 0.02$ Alternative Hypothesis: $H_1: p eq 0.02$ This is a two-tailed test.

• Population Mean: Claim: The mean time to complete an exam is longer than 60 minutes. Null Hypothesis: $H_0: ext{mean} = 60$ Alternative Hypothesis: $H_1: ext{mean} > 60$ This is a right-tailed test.

• Population Standard Deviation: Claim: The standard deviation of detergent bottle contents is less than 0.23 ounces with a new machine. Null Hypothesis: $H_0: ext{standard deviation} = 0.23$ Alternative Hypothesis: $H_1: ext{standard deviation} < 0.23$ This is a left-tailed test.

Probability Example: Coin Flipping

Calculating Probability

Suppose a fair coin is flipped five times, and all outcomes are tails. The probability of this event is:

• Each flip is independent, with $P( ext{Tails}) = rac{1}{2}$.

• Probability of five tails in a row: $P( ext{five tails}) = igg( rac{1}{2}igg)^5 = 0.03125$

• This outcome is possible but unlikely; in 100 sets of 5 flips, about 3 would be all tails.

Interpreting Results

• Either the coin is fair and the result is due to chance, or the coin is biased.

• Hypothesis testing helps decide if the observed result is statistically significant.

Outcomes and Errors in Hypothesis Testing

Possible Outcomes

There are four possible outcomes when conducting a hypothesis test:

1. Reject $H_0$ when $H_1$ is true (Correct Decision)

2. Do not reject $H_0$ when $H_0$ is true (Correct Decision)

3. Reject $H_0$ when $H_0$ is true (Type I Error)

4. Do not reject $H_0$ when $H_1$ is true (Type II Error)

	$H_0$ True	$H_1$ True
Do Not Reject $H_0$	Correct Conclusion	Type II Error
Reject $H_0$	Type I Error	Correct Conclusion

Type I and Type II Errors

• Type I Error ($oldsymbol{ ext{False Positive}}$): Rejecting the null hypothesis when it is actually true. Probability: $oxed{ ext{Type I error probability} = oldsymbol{ ext{alpha}} = oldsymbol{ ext{P(Type I error)}}}$

• Type II Error ($oldsymbol{ ext{False Negative}}$): Not rejecting the null hypothesis when the alternative hypothesis is true. Probability: $oxed{ ext{Type II error probability} = oldsymbol{ ext{beta}} = oldsymbol{ ext{P(Type II error)}}}$

As the probability of a Type I error ($oldsymbol{ ext{alpha}}$) increases, the probability of a Type II error ($oldsymbol{ ext{beta}}$) decreases, and vice versa.

Stating Conclusions in Hypothesis Testing

Interpreting Results

• If sample evidence leads to rejection of $H_0$, conclude there is sufficient evidence to support $H_1$.

• If sample evidence does not lead to rejection of $H_0$, conclude there is not sufficient evidence to support $H_1$.

• Important: We never "accept" the null hypothesis; we only fail to reject it, similar to a court verdict of "not guilty" rather than "innocent."

Summary Table: Hypothesis Test Structure

Test Type	Null Hypothesis ($H_0$)	Alternative Hypothesis ($H_1$)	Symbolic Form
Proportion (Two-Tailed)	The proportion is 0.02	The proportion is not 0.02	$H_0: p = 0.02$ $H_1: p eq 0.02$
Mean (Right-Tailed)	The mean is 60 minutes	The mean is greater than 60 minutes	$H_0: ext{mean} = 60$ $H_1: ext{mean} > 60$
Standard Deviation (Left-Tailed)	The standard deviation is 0.23 ounces	The standard deviation is less than 0.23 ounces	$H_0: ext{sd} = 0.23$ $H_1: ext{sd} < 0.23$

Key Formulas

• Probability of Type I Error: $oxed{ ext{alpha} = P( ext{Type I error}) = P( ext{rejecting } H_0 ext{ when } H_0 ext{ is true})}$

• Probability of Type II Error: $oxed{ ext{beta} = P( ext{Type II error}) = P( ext{not rejecting } H_0 ext{ when } H_1 ext{ is true})}$

Additional info:

• These notes cover the foundational logic and structure of hypothesis testing, including formulation, interpretation, and error analysis, as required for college-level statistics.

• Examples provided illustrate hypothesis tests for proportions, means, and standard deviations, which are central to statistical inference.