Null and Alternative Hypotheses: Parameters and Direction
Translate a research claim into hypotheses about a population parameter, put equality in the null hypothesis, choose the alternative direction before seeing the data, and keep sample statistics out of the hypotheses.
Null and Alternative Hypotheses: Parameters and Direction: direct answer
In a null and alternative hypotheses problem, begin by identifying the population target and the inferential role of the data. Translate a research claim into hypotheses about a population parameter, put equality in the null hypothesis, choose the alternative direction before seeing the data, and keep sample statistics out of the hypotheses.
This page uses null and alternative hypotheses as its single primary search focus. The lesson, numerical cases, and retained questions are restricted to that intent so the page does not function as a generic inference question bank.
Quick reference for Null and Alternative Hypotheses: Parameters and Direction
| Null hypothesis | Contains the equality benchmark |
|---|---|
| Alternative hypothesis | Uses <, >, or ≠ |
| Objects in hypotheses | Population parameters, not sample statistics |
| Direction | Chosen from the question before seeing data |
| Two-group order | Define group 1 minus group 2 explicitly |
| Conclusion | Returns to the alternative claim in context |
Concept mastery: null and alternative hypotheses
Write hypotheses about parameters, not statistics
Hypotheses concern unknown population quantities such as p, mu, p1-p2, or mu1-mu2. A sample statistic such as p-hat or x-bar is observed data and therefore does not belong as the unknown quantity in H0 or Ha.
Put equality in H0
The null hypothesis uses an equality because a significance test needs a specific reference value or relationship to generate its sampling distribution. The alternative uses <, >, or not-equal according to the substantive claim.
Translate words into symbols carefully
“Higher,” “greater,” and “increased” usually imply a greater-than alternative; “lower,” “reduced,” and “less” imply a less-than alternative; “different” or “changed” without direction implies a two-sided alternative.
Define every parameter in context
Before writing symbols, define what p or mu represents, including the population and response. Good parameter definitions prevent ambiguous hypotheses and make the final conclusion much easier to write correctly.
Keep the null benchmark from the claim
If the historical rate is 0.42, H0:p=0.42. If the claim is equality between two populations, H0:p1-p2=0. The sample result should never replace the benchmark just because it is the number you observed.
Do not choose direction after seeing data
The direction belongs to the research question. Looking at p-hat or x-bar first and then choosing the tail that makes the result more significant effectively changes the procedure and invalidates the intended Type I error control.
Distinguish a parameter value from a sample target
A statement such as “at least 60%” needs careful translation depending on the testing convention and the boundary used for H0. In introductory inference, the equality boundary usually supplies the null value and the directional alternative captures the claim under investigation.
Match group order to the verbal claim
For a difference such as p1-p2, “group 1 has a higher proportion” means Ha:p1-p2>0. Reversing the subtraction without changing the sign produces a hypothesis that contradicts the words.
Use two-sided alternatives for genuine nondirectional questions
If departures in either direction matter and no direction was prespecified, Ha uses not-equal. The p-value then includes evidence in both tails of the reference distribution.
Make the final conclusion answer Ha
The conclusion should return to the alternative hypothesis in context. The null supplies the benchmark, but the research question usually asks whether the data provide evidence for the alternative claim.
Boundary values and verbal claims such as “at least”
Claims such as “at least 60%” or “no more than 12 minutes” require careful attention to the boundary that defines the null model. In an introductory significance test, the equality boundary commonly supplies the value used to construct the reference distribution, while the alternative records the directional departure being investigated. The important skill is not to copy the inequality from English mechanically; define the parameter, identify which departure would count as evidence, and make the symbolic hypotheses match that research purpose.
Difference parameters must keep a permanent group order
For p₁−p₂ or μ₁−μ₂, write the group definitions beside the parameter before choosing the alternative. If group 1 is the new program and group 2 is the existing program, a claim that the new program is higher corresponds to a positive alternative. Reversing the subtraction changes the required inequality sign. Keeping the order visible from hypotheses through test statistic and conclusion prevents a common error in which correct arithmetic is attached to the opposite substantive claim.
Worked analysis for null and alternative hypotheses
Hypothesis-writing case 1: Public High School
The study asks whether a completion proportion is higher than 0.60 for seniors who submitted the financial-aid form. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.60; the alternative is Hₐ:p>0.60. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p>0.60, not whether the sample statistic itself satisfies the inequality.
Hypothesis-writing case 2: Municipal Water Office
The study asks whether a mean response time is below 12 minutes for households reporting no service interruption. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=12; the alternative is Hₐ:μ<12. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ<12, not whether the sample statistic itself satisfies the inequality.
Hypothesis-writing case 3: State Park
The study asks whether two population proportions differ for visitors who used the marked trail system. Define p₁−p₂ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p₁−p₂=0; the alternative is Hₐ:p₁−p₂≠0. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p₁−p₂ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p₁−p₂≠0, not whether the sample statistic itself satisfies the inequality.
Hypothesis-writing case 4: Food Cooperative
The study asks whether a mean score exceeds 75 points for members who renewed before the deadline. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=75; the alternative is Hₐ:μ>75. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ>75, not whether the sample statistic itself satisfies the inequality.
Hypothesis-writing case 5: University Advising Center
The study asks whether a defect proportion is less than 0.04 for appointments starting within ten minutes. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.04; the alternative is Hₐ:p<0.04. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p<0.04, not whether the sample statistic itself satisfies the inequality.
Hypothesis-writing case 6: Regional Manufacturer
The study asks whether two population means are different for parts meeting the diameter specification. Define μ₁−μ₂ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ₁−μ₂=0; the alternative is Hₐ:μ₁−μ₂≠0. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ₁−μ₂ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ₁−μ₂≠0, not whether the sample statistic itself satisfies the inequality.
Hypothesis-writing case 7: Public Health Clinic
The study asks whether a renewal proportion exceeds 0.50 for clients returning for the scheduled checkup. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.50; the alternative is Hₐ:p>0.50. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p>0.50, not whether the sample statistic itself satisfies the inequality.
Hypothesis-writing case 8: Urban Recreation Program
The study asks whether a mean wait is not 18 minutes for participants completing the eight-week session. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=18; the alternative is Hₐ:μ≠18. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ≠18, not whether the sample statistic itself satisfies the inequality.
Hypothesis-writing case 9: School District
The study asks whether a completion proportion is higher than 0.60 for families responding to the annual survey. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.60; the alternative is Hₐ:p>0.60. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p>0.60, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 9: School District context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 10: Local Election Office
The study asks whether a mean response time is below 12 minutes for mailed ballots returned before election day. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=12; the alternative is Hₐ:μ<12. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ<12, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 10: Local Election Office context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 11: Energy-Efficiency Pilot
The study asks whether two population proportions differ for homes meeting the target reduction. Define p₁−p₂ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p₁−p₂=0; the alternative is Hₐ:p₁−p₂≠0. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p₁−p₂ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p₁−p₂≠0, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 11: Energy-Efficiency Pilot context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 12: Community Broadband Project
The study asks whether a mean score exceeds 75 points for households achieving the advertised speed. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=75; the alternative is Hₐ:μ>75. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ>75, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 12: Community Broadband Project context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 13: Regional Bus Network
The study asks whether a defect proportion is less than 0.04 for trips arriving within the on-time window. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.04; the alternative is Hₐ:p<0.04. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p<0.04, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 13: Regional Bus Network context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 14: Campus Dining Service
The study asks whether two population means are different for transactions using reusable containers. Define μ₁−μ₂ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ₁−μ₂=0; the alternative is Hₐ:μ₁−μ₂≠0. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ₁−μ₂ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ₁−μ₂≠0, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 14: Campus Dining Service context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 15: Workforce Training Program
The study asks whether a renewal proportion exceeds 0.50 for participants earning the credential. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.50; the alternative is Hₐ:p>0.50. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p>0.50, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 15: Workforce Training Program context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 16: County Recycling Audit
The study asks whether a mean wait is not 18 minutes for sampled loads meeting contamination limits. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=18; the alternative is Hₐ:μ≠18. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ≠18, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 16: County Recycling Audit context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 17: University Residence Halls
The study asks whether a completion proportion is higher than 0.60 for rooms passing the first safety inspection. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.60; the alternative is Hₐ:p>0.60. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p>0.60, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 17: University Residence Halls context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 18: Telehealth Pilot
The study asks whether a mean response time is below 12 minutes for appointments completed without rescheduling. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=12; the alternative is Hₐ:μ<12. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ<12, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 18: Telehealth Pilot context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 19: Public Museum
The study asks whether two population proportions differ for visitors using the audio guide. Define p₁−p₂ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p₁−p₂=0; the alternative is Hₐ:p₁−p₂≠0. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p₁−p₂ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p₁−p₂≠0, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 19: Public Museum context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 20: Youth Sports League
The study asks whether a mean score exceeds 75 points for players completing concussion training. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=75; the alternative is Hₐ:μ>75. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ>75, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 20: Youth Sports League context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 21: Rural Pharmacy Network
The study asks whether a defect proportion is less than 0.04 for prescriptions filled within the service target. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.04; the alternative is Hₐ:p<0.04. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p<0.04, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 21: Rural Pharmacy Network context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 22: City Tree Program
The study asks whether two population means are different for new plantings surviving the first year. Define μ₁−μ₂ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ₁−μ₂=0; the alternative is Hₐ:μ₁−μ₂≠0. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ₁−μ₂ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ₁−μ₂≠0, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 22: City Tree Program context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 23: District Tutoring Initiative
The study asks whether a renewal proportion exceeds 0.50 for students attending at least six sessions. Define p as the corresponding population parameter before writing symbols. The null hypothesis is H₀:p=0.50; the alternative is Hₐ:p>0.50. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace p in the hypotheses. The one-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:p>0.50, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 23: District Tutoring Initiative context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Hypothesis-writing case 24: Regional Call Center
The study asks whether a mean wait is not 18 minutes for calls resolved during the first contact. Define μ as the corresponding population parameter before writing symbols. The null hypothesis is H₀:μ=18; the alternative is Hₐ:μ≠18. Equality stays in H₀ because the null model needs a specific boundary or benchmark from which a reference distribution can be constructed.
The symbols refer to the population, so an observed sample proportion or sample mean must not replace μ in the hypotheses. The two-sided direction comes from the wording of the research question and should be set before examining the sample result. A final test conclusion would answer whether the data provide convincing evidence for Hₐ:μ≠18, not whether the sample statistic itself satisfies the inequality. In the Hypothesis-writing case 24: Regional Call Center context, this checkpoint must be tied to the stated population and data structure before the numerical conclusion is accepted.
Null And Alternative Hypotheses multiple-choice practice
Question 1. Write H₀ and Hₐ
Community Health Network asks whether the population proportion for follow-up completion rate is greater than 0.62. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: C
Hypotheses describe population states, so p must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0.62; the phrase “greater than 0.62” determines Hₐ:p>0.62. This direction should be fixed from the question before any sample result is inspected. In the Community Health Network setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 2. Write H₀ and Hₐ
Regional College asks whether the population mean for mean weekly study time is less than 18. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: B
Hypotheses describe population states, so μ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 18; the phrase “less than 18” determines Hₐ:μ<18. This direction should be fixed from the question before any sample result is inspected. In the Regional College setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 3. Write H₀ and Hₐ
City Transit Agency asks whether the population difference in proportions for on-time arrival proportion is different from 0. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: A
Hypotheses describe population states, so p₁−p₂ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0; the phrase “different from 0” determines Hₐ:p₁−p₂≠0. This direction should be fixed from the question before any sample result is inspected. In the City Transit Agency setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 4. Write H₀ and Hₐ
County Library System asks whether the population difference in means for program participation rate is greater than 0. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: D
Hypotheses describe population states, so μ₁−μ₂ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0; the phrase “greater than 0” determines Hₐ:μ₁−μ₂>0. This direction should be fixed from the question before any sample result is inspected. In the County Library System setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 5. Write H₀ and Hₐ
Manufacturing Plant asks whether the population proportion for mean component strength is greater than 0.62. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: C
Hypotheses describe population states, so p must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0.62; the phrase “greater than 0.62” determines Hₐ:p>0.62. This direction should be fixed from the question before any sample result is inspected. In the Manufacturing Plant setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 6. Write H₀ and Hₐ
School District asks whether the population mean for graduation-plan completion proportion is less than 18. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: B
Hypotheses describe population states, so μ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 18; the phrase “less than 18” determines Hₐ:μ<18. This direction should be fixed from the question before any sample result is inspected. In the School District setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 7. Write H₀ and Hₐ
Public Parks Department asks whether the population difference in proportions for visitor satisfaction proportion is different from 0. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: A
Hypotheses describe population states, so p₁−p₂ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0; the phrase “different from 0” determines Hₐ:p₁−p₂≠0. This direction should be fixed from the question before any sample result is inspected. In the Public Parks Department setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 8. Write H₀ and Hₐ
Energy Pilot asks whether the population difference in means for mean household savings is greater than 0. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: D
Hypotheses describe population states, so μ₁−μ₂ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0; the phrase “greater than 0” determines Hₐ:μ₁−μ₂>0. This direction should be fixed from the question before any sample result is inspected. In the Energy Pilot setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 9. Write H₀ and Hₐ
Hospital System asks whether the population proportion for 30-day readmission proportion is greater than 0.62. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: C
Hypotheses describe population states, so p must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0.62; the phrase “greater than 0.62” determines Hₐ:p>0.62. This direction should be fixed from the question before any sample result is inspected. In the Hospital System setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 10. Write H₀ and Hₐ
Workforce Program asks whether the population mean for credential completion rate is less than 18. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: B
Hypotheses describe population states, so μ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 18; the phrase “less than 18” determines Hₐ:μ<18. This direction should be fixed from the question before any sample result is inspected. In the Workforce Program setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 11. Write H₀ and Hₐ
Municipal Call Center asks whether the population difference in proportions for mean resolution time is different from 0. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: A
Hypotheses describe population states, so p₁−p₂ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0; the phrase “different from 0” determines Hₐ:p₁−p₂≠0. This direction should be fixed from the question before any sample result is inspected. In the Municipal Call Center setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 12. Write H₀ and Hₐ
Broadband Project asks whether the population difference in means for household reliability proportion is greater than 0. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: D
Hypotheses describe population states, so μ₁−μ₂ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0; the phrase “greater than 0” determines Hₐ:μ₁−μ₂>0. This direction should be fixed from the question before any sample result is inspected. In the Broadband Project setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 13. Write H₀ and Hₐ
University Housing asks whether the population proportion for mean repair turnaround time is greater than 0.62. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: C
Hypotheses describe population states, so p must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0.62; the phrase “greater than 0.62” determines Hₐ:p>0.62. This direction should be fixed from the question before any sample result is inspected. In the University Housing setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 14. Write H₀ and Hₐ
Food Cooperative asks whether the population mean for member renewal proportion is less than 18. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: B
Hypotheses describe population states, so μ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 18; the phrase “less than 18” determines Hₐ:μ<18. This direction should be fixed from the question before any sample result is inspected. In the Food Cooperative setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 15. Write H₀ and Hₐ
Recreation Department asks whether the population difference in proportions for mean weekly attendance is different from 0. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: A
Hypotheses describe population states, so p₁−p₂ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0; the phrase “different from 0” determines Hₐ:p₁−p₂≠0. This direction should be fixed from the question before any sample result is inspected. In the Recreation Department setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Question 16. Write H₀ and Hₐ
Water Authority asks whether the population difference in means for compliance proportion is greater than 0. Which hypothesis pair is written at the population level and uses the research direction correctly?
Answer: D
Hypotheses describe population states, so μ₁−μ₂ must be defined for the population rather than replaced by a sample statistic. Equality belongs in H₀ at the benchmark 0; the phrase “greater than 0” determines Hₐ:μ₁−μ₂>0. This direction should be fixed from the question before any sample result is inspected. In the Water Authority setting, the parameter definition must name the population and response before the symbolic pair is interpreted.
Null And Alternative Hypotheses free-response practice
FRQ 1. Translate the research claim
Manufacturing Plant is planning a study of mean component strength. The research claim is that the relevant population difference in proportions is different from 0. Define the parameter in context, write H₀ and Hₐ, explain why equality belongs in H₀, identify whether the alternative is one-sided or two-sided, and explain what would be wrong with choosing the direction after seeing the sample result.
Model response
Define p₁−p₂ as the population difference in proportions corresponding to mean component strength for the population of interest. Write H₀:p₁−p₂=0 and Hₐ:p₁−p₂≠0. Equality supplies the benchmark needed to build the null reference distribution. Because the claim is “different from 0,” the alternative is two-sided. Choosing the direction after observing the sample would use the data twice and distort the intended Type I error control. The hypotheses must be sensible before p-hat, x-bar, or any test statistic is known.
FRQ 2. Translate the research claim
Energy Pilot is planning a study of mean household savings. The research claim is that the relevant population difference in means is greater than 0. Define the parameter in context, write H₀ and Hₐ, explain why equality belongs in H₀, identify whether the alternative is one-sided or two-sided, and explain what would be wrong with choosing the direction after seeing the sample result.
Model response
Define μ₁−μ₂ as the population difference in means corresponding to mean household savings for the population of interest. Write H₀:μ₁−μ₂=0 and Hₐ:μ₁−μ₂>0. Equality supplies the benchmark needed to build the null reference distribution. Because the claim is “greater than 0,” the alternative is one-sided. Choosing the direction after observing the sample would use the data twice and distort the intended Type I error control. The hypotheses must be sensible before p-hat, x-bar, or any test statistic is known.
FRQ 3. Translate the research claim
Municipal Call Center is planning a study of mean resolution time. The research claim is that the relevant population proportion is greater than 0.62. Define the parameter in context, write H₀ and Hₐ, explain why equality belongs in H₀, identify whether the alternative is one-sided or two-sided, and explain what would be wrong with choosing the direction after seeing the sample result.
Model response
Define p as the population proportion corresponding to mean resolution time for the population of interest. Write H₀:p=0.62 and Hₐ:p>0.62. Equality supplies the benchmark needed to build the null reference distribution. Because the claim is “greater than 0.62,” the alternative is one-sided. Choosing the direction after observing the sample would use the data twice and distort the intended Type I error control. The hypotheses must be sensible before p-hat, x-bar, or any test statistic is known.
FRQ 4. Translate the research claim
Food Cooperative is planning a study of member renewal proportion. The research claim is that the relevant population mean is less than 18. Define the parameter in context, write H₀ and Hₐ, explain why equality belongs in H₀, identify whether the alternative is one-sided or two-sided, and explain what would be wrong with choosing the direction after seeing the sample result.
Model response
Define μ as the population mean corresponding to member renewal proportion for the population of interest. Write H₀:μ=18 and Hₐ:μ<18. Equality supplies the benchmark needed to build the null reference distribution. Because the claim is “less than 18,” the alternative is one-sided. Choosing the direction after observing the sample would use the data twice and distort the intended Type I error control. The hypotheses must be sensible before p-hat, x-bar, or any test statistic is known.
FRQ 5. Translate the research claim
Community Health Network is planning a study of follow-up completion rate. The research claim is that the relevant population difference in proportions is different from 0. Define the parameter in context, write H₀ and Hₐ, explain why equality belongs in H₀, identify whether the alternative is one-sided or two-sided, and explain what would be wrong with choosing the direction after seeing the sample result.
Model response
Define p₁−p₂ as the population difference in proportions corresponding to follow-up completion rate for the population of interest. Write H₀:p₁−p₂=0 and Hₐ:p₁−p₂≠0. Equality supplies the benchmark needed to build the null reference distribution. Because the claim is “different from 0,” the alternative is two-sided. Choosing the direction after observing the sample would use the data twice and distort the intended Type I error control. The hypotheses must be sensible before p-hat, x-bar, or any test statistic is known.
FRQ 6. Translate the research claim
County Library System is planning a study of program participation rate. The research claim is that the relevant population difference in means is greater than 0. Define the parameter in context, write H₀ and Hₐ, explain why equality belongs in H₀, identify whether the alternative is one-sided or two-sided, and explain what would be wrong with choosing the direction after seeing the sample result.
Model response
Define μ₁−μ₂ as the population difference in means corresponding to program participation rate for the population of interest. Write H₀:μ₁−μ₂=0 and Hₐ:μ₁−μ₂>0. Equality supplies the benchmark needed to build the null reference distribution. Because the claim is “greater than 0,” the alternative is one-sided. Choosing the direction after observing the sample would use the data twice and distort the intended Type I error control. The hypotheses must be sensible before p-hat, x-bar, or any test statistic is known.
Next steps after mastering null and alternative hypotheses
Practice hypothesis writing without calculating anything. Convert ten verbal claims into parameter definitions and symbolic H₀/Hₐ pairs, then reverse the group order on difference problems to verify that the inequality sign must reverse as well. This isolates the logic of the research question from the arithmetic of a later test.
When reviewing errors, label each one as a parameter error, equality-placement error, direction error, or data-peeking error. A correct pair of hypotheses should remain sensible even before the sample statistics are revealed, because the hypotheses describe population states rather than observed results.
