Inference (Hypothesis Test or Confidence Interval)

Carry out a one-proportion z-test: AP Stats FRQ practice

3 versions of one AP Statistics free-response question, each with a 10-point rubric and a model answer for every point. Write your response on paper before you open the rubric.

These are StatsLearn originals written in the AP style, not released exam questions. Topics: 3.5 Setting Up a Test for a Population Proportion, 3.6 p-Values, 3.7 Carrying Out a Test for a Population Proportion.

Version 1

A transit agency claims that 42% of riders use a mobile ticket. A random sample of 180 riders from a population of 12000 includes 88 mobile-ticket users. Test whether the true proportion is greater than the claimed value at alpha = 0.05.

  1. (a) Define the parameter and state hypotheses.
  2. (b) Check the conditions for a one-proportion z-test.
  3. (c) Calculate the sample proportion, standard error, test statistic, and p-value.
  4. (d) State the decision and conclusion in context.
Show the rubric and model answers

Part (a): 2 points

  • Defines the population parameter. Let p be the true proportion of all agency riders who use a mobile ticket.
  • States the null and directional alternative hypotheses. H0: p = 0.42. Ha: p is greater than 0.42.

Part (b): 3 points

  • Checks the random condition. The problem states that the riders were randomly sampled, so the random condition is met.
  • Checks the 10% condition. 180 is no more than 10% of 12000, so treating observations as independent is reasonable.
  • Checks large counts under the null. Under H0, expected counts are 75.6 successes and 104.4 failures, both at least 10.

Part (c): 4 points

  • Computes the sample proportion. p-hat = 88/180 = 0.4889.
  • Computes the null standard error. SE0 = sqrt(0.42(0.58)/180) = 0.0368.
  • Computes the z statistic. z = (0.4889 - 0.42)/0.0368 = 1.873.
  • Computes the directional p-value. The right-tailed p-value is 0.0306.

Part (d): 1 point

  • Makes the decision and conclusion in context. Reject H0. There is convincing evidence that the true mobile-ticket proportion is greater than 0.42.

Version 2

A transit agency claims that 46% of riders use a mobile ticket. A random sample of 260 riders from a population of 14000 includes 136 mobile-ticket users. Test whether the true proportion is greater than the claimed value at alpha = 0.05.

  1. (a) Define the parameter and state hypotheses.
  2. (b) Check the conditions for a one-proportion z-test.
  3. (c) Calculate the sample proportion, standard error, test statistic, and p-value.
  4. (d) State the decision and conclusion in context.
Show the rubric and model answers

Part (a): 2 points

  • Defines the population parameter. Let p be the true proportion of all agency riders who use a mobile ticket.
  • States the null and directional alternative hypotheses. H0: p = 0.46. Ha: p is greater than 0.46.

Part (b): 3 points

  • Checks the random condition. The problem states that the riders were randomly sampled, so the random condition is met.
  • Checks the 10% condition. 260 is no more than 10% of 14000, so treating observations as independent is reasonable.
  • Checks large counts under the null. Under H0, expected counts are 119.6 successes and 140.4 failures, both at least 10.

Part (c): 4 points

  • Computes the sample proportion. p-hat = 136/260 = 0.5231.
  • Computes the null standard error. SE0 = sqrt(0.46(0.54)/260) = 0.0309.
  • Computes the z statistic. z = (0.5231 - 0.46)/0.0309 = 2.041.
  • Computes the directional p-value. The right-tailed p-value is 0.0206.

Part (d): 1 point

  • Makes the decision and conclusion in context. Reject H0. There is convincing evidence that the true mobile-ticket proportion is greater than 0.46.

Version 3

A transit agency claims that 50% of riders use a mobile ticket. A random sample of 340 riders from a population of 16000 includes 190 mobile-ticket users. Test whether the true proportion is greater than the claimed value at alpha = 0.05.

  1. (a) Define the parameter and state hypotheses.
  2. (b) Check the conditions for a one-proportion z-test.
  3. (c) Calculate the sample proportion, standard error, test statistic, and p-value.
  4. (d) State the decision and conclusion in context.
Show the rubric and model answers

Part (a): 2 points

  • Defines the population parameter. Let p be the true proportion of all agency riders who use a mobile ticket.
  • States the null and directional alternative hypotheses. H0: p = 0.50. Ha: p is greater than 0.50.

Part (b): 3 points

  • Checks the random condition. The problem states that the riders were randomly sampled, so the random condition is met.
  • Checks the 10% condition. 340 is no more than 10% of 16000, so treating observations as independent is reasonable.
  • Checks large counts under the null. Under H0, expected counts are 170.0 successes and 170.0 failures, both at least 10.

Part (c): 4 points

  • Computes the sample proportion. p-hat = 190/340 = 0.5588.
  • Computes the null standard error. SE0 = sqrt(0.50(0.50)/340) = 0.0271.
  • Computes the z statistic. z = (0.5588 - 0.50)/0.0271 = 2.169.
  • Computes the directional p-value. The right-tailed p-value is 0.0150.

Part (d): 1 point

  • Makes the decision and conclusion in context. Reject H0. There is convincing evidence that the true mobile-ticket proportion is greater than 0.50.

Keep going