Resources

Consider a one-sample zz-test of H0:μ=μ0H_0: \mu = \mu_0 versus Ha:μ>μ0H_a: \mu > \mu_0 with fixed true mean discrepancy Δ=μ−μ0=0.05 mm\Delta = \mu - \mu_0 = 0.05\text{ mm}, known population standard deviation σ=1.00 mm\sigma = 1.00\text{ mm}, and significance level α=0.05\alpha = 0.05 (zcrit=1.645z_{\text{crit}} = 1.645). (a) Express the test statistic z(n)z(n) and Cohen's dd as functions of the sample size nn. (b) Determine lim⁡n→∞z(n)\lim_{n \to \infty} z(n), the resulting limit of the pp-value as n→∞n \to \infty, and the limit of Cohen's dd as n→∞n \to \infty. (c) Explain the asymptotic behavior of statistical hypothesis testing: why does any trivially small, non-zero difference inevitably become statistically significant as n→∞n \to \infty, and why does this necessitate reporting effect size point estimates and confidence intervals for practical interpretation?

Answer and steps

Final answer

(a) z(n)=Δσn=0.051.00n=0.05nz(n) = \frac{\Delta}{\frac{\sigma}{\sqrt{n}}} = \frac{0.05}{\frac{1.00}{\sqrt{n}}} = 0.05\sqrt{n}; Cohen's d=Δσ=0.051.00=0.050d = \frac{\Delta}{\sigma} = \frac{0.05}{1.00} = 0.050 (independent of nn). (b) lim⁡n→∞z(n)=lim⁡n→∞(0.05n)=∞\lim_{n \to \infty} z(n) = \lim_{n \to \infty} (0.05\sqrt{n}) = \infty; consequently, lim⁡n→∞p-value=lim⁡z→∞[1−Φ(z)]=0\lim_{n \to \infty} p\text{-value} = \lim_{z \to \infty} [1 - \Phi(z)] = 0; while lim⁡n→∞d=0.050\lim_{n \to \infty} d = 0.050. (c) As n→∞n \to \infty, the standard error σn→0\frac{\sigma}{\sqrt{n}} \to 0, driving the test statistic to infinity and the pp-value to zero for any non-zero difference Δ\Delta, no matter how microscopic. Therefore, statistical significance alone cannot assess whether an effect matters; reporting effect sizes (such as Cohen's d=0.050d = 0.050) and confidence intervals is mandatory to evaluate practical significance.

Steps

Part (a): Expressing z(n)z(n) and Cohen's dd

The one-sample zz-test statistic tests the null hypothesis H0:μ=μ0H_0: \mu = \mu_0 against Ha:μ>μ0H_a: \mu > \mu_0. When the true population mean is μ=μ0+Δ\mu = \mu_0 + \Delta, the expected value of the sample mean is E[Xˉ]=μ0+Δ\text{E}[\bar{X}] = \mu_0 + \Delta.

The expected value of the test statistic as a function of sample size nn is given by: z(n)=Xˉ−μ0σn=Δσn=Δnσz(n) = \frac{\bar{X} - \mu_0}{\frac{\sigma}{\sqrt{n}}} = \frac{\Delta}{\frac{\sigma}{\sqrt{n}}} = \frac{\Delta \sqrt{n}}{\sigma}

Substituting the given values Δ=0.05 mm\Delta = 0.05\text{ mm} and σ=1.00 mm\sigma = 1.00\text{ mm}: z(n)=0.051.00n=0.05nz(n) = \frac{0.05}{\frac{1.00}{\sqrt{n}}} = 0.05\sqrt{n}

Cohen's dd is a standardized measure of effect size that quantifies the magnitude of the difference relative to the population standard deviation: d=μ−μ0σ=Δσ=0.051.00=0.050d = \frac{\mu - \mu_0}{\sigma} = \frac{\Delta}{\sigma} = \frac{0.05}{1.00} = 0.050

Because Cohen's dd depends solely on the population parameters Δ\Delta and σ\sigma, it is completely independent of the sample size nn.


Part (b): Evaluating the Limits as n→∞n \to \infty

  1. Limit of the test statistic z(n)z(n): lim⁡n→∞z(n)=lim⁡n→∞(0.05n)=∞\lim_{n \to \infty} z(n) = \lim_{n \to \infty} (0.05\sqrt{n}) = \infty
  2. Limit of the pp-value: For an upper-tailed test, the pp-value corresponds to the probability under the standard normal distribution Φ\Phi in the upper tail: p-value=P(Z≥z(n))=1−Φ(z(n))p\text{-value} = P(Z \ge z(n)) = 1 - \Phi(z(n)) Taking the limit as z(n)→∞z(n) \to \infty: lim⁡n→∞p-value=lim⁡z→∞[1−Φ(z)]=1−1=0\lim_{n \to \infty} p\text{-value} = \lim_{z \to \infty} [1 - \Phi(z)] = 1 - 1 = 0
  3. Limit of Cohen's dd: Since Cohen's dd is constant with respect to nn: lim⁡n→∞d=0.050\lim_{n \to \infty} d = 0.050

Part (c): Statistical vs. Practical Significance

As n→∞n \to \infty, the standard error of the mean, SE=σn\text{SE} = \frac{\sigma}{\sqrt{n}}, approaches zero: lim⁡n→∞σn=0\lim_{n \to \infty} \frac{\sigma}{\sqrt{n}} = 0

Because the standard error shrinks toward zero, any non-zero difference Δ>0\Delta > 0—no matter how minuscule or trivial—is divided by an arbitrarily small standard error. This forces the test statistic z(n)z(n) toward infinity and drives the pp-value toward zero (p<α=0.05p < \alpha = 0.05). Consequently, with a sufficiently large sample size, any microscopic discrepancy is virtually guaranteed to be rejected as "statistically significant."

However, statistical significance only answers the binary question of whether the observed data are consistent with the null hypothesis (Δ=0\Delta = 0). It provides no information about the practical importance or magnitude of that difference. Here, Cohen's d=0.050d = 0.050 represents an extremely negligible effect (conventionally, d=0.2d = 0.2 is considered small, d=0.5d = 0.5 medium, and d=0.8d = 0.8 large).

Therefore, reporting effect sizes and confidence intervals alongside pp-values is essential:

  • Effect sizes quantify the magnitude of the real-world impact independently of sample size.
  • Confidence intervals bound the true parameter value in real-world units, allowing researchers to see whether the entire plausible range of values lies within a region of practical equivalence or negligible clinical/engineering importance.

Typical errors

OverviewACEpath Teacher's Diagnosis

01 The student confuses the standard error with the population standard deviation in the denominator, missing the factor of √n in the test statistic.

Sample Student Work

Reviewed by ACEpath Teacher
Step-by-Step Review
Walk Through It with Your Teacher
Incorrect
  1. Step 1

    z(n)=μ−μ0σ=0.051.00=0.05z(n) = \frac{\mu - \mu_0}{\sigma} = \frac{0.05}{1.00} = 0.05

    Correct

    z(n)=μ−μ0σn=0.051.00n=0.05nz(n) = \frac{\mu - \mu_0}{\frac{\sigma}{\sqrt{n}}} = \frac{0.05}{\frac{1.00}{\sqrt{n}}} = 0.05\sqrt{n}

    A sample mean Xˉ\bar{X} has standard error SE=σn\text{SE} = \frac{\sigma}{\sqrt{n}}, not σ\sigma. Dividing by σ\sigma computes a single observation's zz-score rather than the test statistic for a sample mean of size nn.

  2. Step 2

    Cohen’s d=Δσ=0.051.00=0.05\text{Cohen's } d = \frac{\Delta}{\sigma} = \frac{0.05}{1.00} = 0.05

  3. Step 3

    lim⁡n→∞z(n)=0.05  ⟹  lim⁡n→∞p-value=P(Z≥0.05)=0.4801\lim_{n \to \infty} z(n) = 0.05 \implies \lim_{n \to \infty} p\text{-value} = P(Z \ge 0.05) = 0.4801

    Correct

    lim⁡n→∞z(n)=lim⁡n→∞0.05n=∞  ⟹  lim⁡n→∞p-value=0\lim_{n \to \infty} z(n) = \lim_{n \to \infty} 0.05\sqrt{n} = \infty \implies \lim_{n \to \infty} p\text{-value} = 0

    Because your z(n)z(n) omitted n\sqrt{n}, your limits were held constant at 0.050.05 rather than diverging to ∞\infty for the test statistic and converging to 00 for the pp-value.

  4. Step 4

    lim⁡n→∞Cohen’s d=0.05\lim_{n \to \infty} \text{Cohen's } d = 0.05

  5. Step 5

    As n→∞n \to \infty, the test statistic stays at 0.050.05 and does not exceed 1.6451.645, so H0H_0 is never rejected; effect sizes and CIs are needed because test statistics cannot detect small differences.

    Correct

    As n→∞n \to \infty, SE→0\text{SE} \to 0, driving z(n)→∞z(n) \to \infty and p→0p \to 0. Any non-zero difference inevitably becomes statistically significant; effect sizes and CIs are needed because pp-values cannot assess practical importance.

    Large samples do not fail to detect small differences; rather, they overpower tests so that even trivial differences inevitably become statistically significant (p<αp < \alpha). Effect sizes and confidence intervals are needed to judge whether a detectable difference is practically meaningful.

WHAT YOU DID WELL
Compliments from Your Teacher

You correctly computed Cohen's d=0.050d = 0.050 from the population parameters.

You correctly identified that Cohen's dd is independent of sample size nn.

You accurately computed the upper-tail probability for your stated zz-value.

THE ERROR
Where Your ACEpath Teacher Spotted the Mistake
Confusing standard error with standard deviation

no

WHY IT HAPPENED
ACEpath Teacher's Diagnosis

You confused the standard deviation of individual observations (σ\sigma) with the standard error of the sample mean (σ/n\sigma/\sqrt{n}) in the denominator of the test statistic. This caused you to miss the n\sqrt{n} scaling factor, leading to the inverted misconception that test statistics cannot detect small differences with large samples.

REASONING PATH
Your ACEpath teacher’s suggested approach
Denominator of a Sample Mean Test Statistic
  • SE(Xˉ)=σn\text{SE}(\bar{X}) = \frac{\sigma}{\sqrt{n}}, so z=Xˉ−μ0σ/n=(Xˉ−μ0)nσz = \frac{\bar{X} - \mu_0}{\sigma / \sqrt{n}} = \frac{(\bar{X} - \mu_0)\sqrt{n}}{\sigma}.

  • Always divide the mean difference by the standard error σn\frac{\sigma}{\sqrt{n}}, ensuring the test statistic scales with n\sqrt{n}.

Large-Sample Asymptotic Behavior
  • For any fixed Δ≠0\Delta \ne 0, lim⁡n→∞z(n)=±∞\lim_{n \to \infty} z(n) = \pm \infty and lim⁡n→∞p-value=0\lim_{n \to \infty} p\text{-value} = 0.

  • Remember that statistical significance tests whether Δ≠0\Delta \ne 0, not whether Δ\Delta is large; sufficiently large samples will detect any non-zero difference.

CORRECTED PATH
ACEpath Teacher's Demonstration
  1. Express the test statistic with standard error in the denominator: z(n)=Δσ/n=0.051.00/n=0.05nz(n) = \frac{\Delta}{\sigma / \sqrt{n}} = \frac{0.05}{1.00 / \sqrt{n}} = 0.05\sqrt{n}.

  2. Evaluate the asymptotic limits: lim⁡n→∞z(n)=lim⁡n→∞0.05n=∞\lim_{n \to \infty} z(n) = \lim_{n \to \infty} 0.05\sqrt{n} = \infty, which yields lim⁡n→∞p-value=lim⁡z→∞[1−Φ(z)]=0\lim_{n \to \infty} p\text{-value} = \lim_{z \to \infty} [1 - \Phi(z)] = 0.

  3. Explain practical significance: because SE→0\text{SE} \to 0, z(n)z(n) inevitably exceeds any critical value (p<0.05p < 0.05) for any Δ>0\Delta > 0, making trivial effects statistically significant. Reporting effect size (d=0.050d = 0.050) and confidence intervals reveals whether the detected effect has real-world importance.

Final answer

(a) z(n)=0.05nz(n) = 0.05\sqrt{n}; Cohen's d=0.050d = 0.050. (b) lim⁡n→∞z(n)=∞\lim_{n \to \infty} z(n) = \infty, lim⁡n→∞p-value=0\lim_{n \to \infty} p\text{-value} = 0, lim⁡n→∞d=0.050\lim_{n \to \infty} d = 0.050. (c) As n→∞n \to \infty, the standard error approaches 00, driving z(n)→∞z(n) \to \infty and p→0p \to 0 for any non-zero difference; effect sizes and confidence intervals are essential to assess practical importance.

NEXT STEPS
Final Takeaways from Your Teacher
  • Review the Central Limit Theorem and note that the sampling distribution of Xˉ\bar{X} has standard deviation σn\frac{\sigma}{\sqrt{n}}.

  • Contrast statistical significance (is the difference distinguishable from zero?) with practical significance (does the magnitude of difference matter?).

OverviewACEpath Teacher's Diagnosis

02 The student incorrectly scales Cohen's d by √n, conflating standardized effect size with the test statistic.

Sample Student Work

Reviewed by ACEpath Teacher
Step-by-Step Review
Walk Through It with Your Teacher
Partially correct
  1. Step 1

    z(n)=Δnσ=0.05nz(n) = \frac{\Delta \sqrt{n}}{\sigma} = 0.05\sqrt{n}

  2. Step 2

    Cohen’s d=Δσn=0.05n\text{Cohen's } d = \frac{\Delta}{\sigma}\sqrt{n} = 0.05\sqrt{n}

    Correct

    d = Δσ\frac{\Delta}{\sigma} = 0.051.00\frac{0.05}{1.00} = 0.050

    Cohen's dd is a population effect size metric defined as μ−μ0σ=Δσ\frac{\mu - \mu_0}{\sigma} = \frac{\Delta}{\sigma}, which does not depend on the sample size nn. You incorrectly multiplied it by n\sqrt{n}, conflating the standardized effect size with the test statistic.

  3. Step 3

    lim⁡n→∞z(n)=∞  ⟹  lim⁡n→∞p-value=0\lim_{n \to \infty} z(n) = \infty \implies \lim_{n \to \infty} p\text{-value} = 0

  4. Step 4

    lim⁡n→∞Cohen’s d=∞\lim_{n \to \infty} \text{Cohen's } d = \infty

    Correct

    lim⁡n→∞d=0.050\lim_{n \to \infty} d = 0.050

    Because you defined Cohen's dd as a function of n\sqrt{n} in Step 2, your limit evaluated to ∞\infty instead of remaining constant at 0.0500.050.

  5. Step 5

    As nn grows, both the test statistic and effect size become infinitely large, which forces the pp-value to zero. Confidence intervals are needed because infinite effect sizes in large samples require bounds on practical significance.

    Correct

    As n→∞n \to \infty, the standard error σn→0\frac{\sigma}{\sqrt{n}} \to 0, forcing z→∞z \to \infty and p→0p \to 0 for any Δ>0\Delta > 0, even when the true difference is negligible. Because statistical significance only indicates p<αp < \alpha, reporting effect sizes (here d=0.050d = 0.050) and confidence intervals is necessary to assess whether the magnitude of the difference has real-world practical importance.

    Because you assumed effect size increases with sample size, your explanation missed the fundamental contrast between statistical significance (which depends heavily on nn) and practical significance (which measures the true, sample-size-invariant effect size).

WHAT YOU DID WELL
Compliments from Your Teacher

You correctly derived the formula for the test statistic z(n)=0.05nz(n) = 0.05\sqrt{n}.

You correctly evaluated the asymptotic limits of the test statistic and the pp-value as n→∞n \to \infty.

THE ERROR
Where Your ACEpath Teacher Spotted the Mistake
Conflating effect size with test statistic scaling

Multiplying Cohen's dd by n\sqrt{n}.

WHY IT HAPPENED
ACEpath Teacher's Diagnosis

Conflating sample-dependent test statistics with population effect sizes. The test statistic z=Xˉ−μ0SEz = \frac{\bar{X} - \mu_0}{\text{SE}} scales with n\sqrt{n} because the standard error SE=σn\text{SE} = \frac{\sigma}{\sqrt{n}} shrinks as sample size grows. In contrast, Cohen's d=μ−μ0σd = \frac{\mu - \mu_0}{\sigma} divides the mean difference by the population standard deviation, making it purely a property of the population distributions and completely invariant to nn.

REASONING PATH
Your ACEpath teacher’s suggested approach
Effect Size vs. Test Statistic Relationship
  • Test Statistic=Effect Size×n\text{Test Statistic} = \text{Effect Size} \times \sqrt{n} (e.g., z=dnz = d\sqrt{n})

  • Keep Cohen's dd defined strictly as Δσ\frac{\Delta}{\sigma}; do not include sample size nn in effect size definitions.

Practical Significance Interpretation
  • Large sample sizes can make minuscule effects statistically significant (p<αp < \alpha) even when practical effect size dd is near zero.

  • Use effect sizes (dd) and confidence intervals to quantify the real-world magnitude of a difference independently of sample size.

CORRECTED PATH
ACEpath Teacher's Demonstration
  1. State the definition of Cohen's dd: d=Δσ=0.051.00=0.050d = \frac{\Delta}{\sigma} = \frac{0.05}{1.00} = 0.050.

  2. Note that dd does not contain nn, so lim⁡n→∞d=0.050\lim_{n \to \infty} d = 0.050.

  3. Explain that because SE=σn→0\text{SE} = \frac{\sigma}{\sqrt{n}} \to 0, z(n)→∞z(n) \to \infty and p→0p \to 0 for any non-zero Δ\Delta. However, d=0.050d = 0.050 remains tiny, demonstrating that statistical significance does not imply a meaningful real-world effect.

Final answer

(a) z(n)=0.05nz(n) = 0.05\sqrt{n}; Cohen's d=0.050d = 0.050. (b) lim⁡n→∞z(n)=∞\lim_{n \to \infty} z(n) = \infty, lim⁡n→∞p-value=0\lim_{n \to \infty} p\text{-value} = 0, lim⁡n→∞d=0.050\lim_{n \to \infty} d = 0.050. (c) Shrinking standard error forces statistical significance (p→0p \to 0) even for trivial differences; effect sizes and confidence intervals reveal whether the true effect has practical importance.

NEXT STEPS
Final Takeaways from Your Teacher
  • Review the fundamental formula z=dnz = d\sqrt{n} to see explicitly how sample size scales the test statistic, not the effect size.

  • Practice interpreting study results where p<0.001p < 0.001 but Cohen's d<0.10d < 0.10 to build intuition for why large datasets require effect size reporting.

OverviewACEpath Teacher's Diagnosis

03 The student divides by √n instead of multiplying by √n when simplifying the complex fraction for z(n).

Sample Student Work

Reviewed by ACEpath Teacher
Step-by-Step Review
Walk Through It with Your Teacher
Incorrect
  1. Step 1

    z(n)=Δσ/n=0.05nz(n) = \frac{\Delta}{\sigma/\sqrt{n}} = \frac{0.05}{\sqrt{n}} and Cohen’s d=Δσ=0.05\text{Cohen's } d = \frac{\Delta}{\sigma} = 0.05

    Correct

    z(n)=Δσn=0.051.00n=0.05nz(n) = \frac{\Delta}{\frac{\sigma}{\sqrt{n}}} = \frac{0.05}{\frac{1.00}{\sqrt{n}}} = 0.05\sqrt{n} and Cohen’s d=0.051.00=0.050\text{Cohen's } d = \frac{0.05}{1.00} = 0.050

    You correctly set up the formula for z(n)z(n) and computed Cohen's d=0.05d = 0.05. However, dividing by the standard error σn\frac{\sigma}{\sqrt{n}} means multiplying by the reciprocal: 0.051n=0.05n\frac{0.05}{\frac{1}{\sqrt{n}}} = 0.05\sqrt{n}, not 0.05n\frac{0.05}{\sqrt{n}}.

  2. Step 2

    lim⁡n→∞z(n)=lim⁡n→∞0.05n=0  ⟹  lim⁡n→∞p-value=P(Z≥0)=0.5  ⟹  lim⁡n→∞Cohen’s d=0.05\lim_{n \to \infty} z(n) = \lim_{n \to \infty} \frac{0.05}{\sqrt{n}} = 0 \implies \lim_{n \to \infty} p\text{-value} = P(Z \ge 0) = 0.5 \implies \lim_{n \to \infty} \text{Cohen's } d = 0.05

    Correct

    lim⁡n→∞z(n)=lim⁡n→∞(0.05n)=∞  ⟹  lim⁡n→∞p-value=0  ⟹  lim⁡n→∞d=0.050\lim_{n \to \infty} z(n) = \lim_{n \to \infty} (0.05\sqrt{n}) = \infty \implies \lim_{n \to \infty} p\text{-value} = 0 \implies \lim_{n \to \infty} d = 0.050

    Your limit evaluation logically followed your expression from Part (a), and your limit for Cohen's dd is correct. Because z(n)=0.05nz(n) = 0.05\sqrt{n} actually grows without bound as n→∞n \to \infty, the pp-value approaches 00, not 0.50.5.

  3. Step 3

    "As nn approaches infinity, the test statistic shrinks to 0, which prevents the test from ever reaching statistical significance. Reporting confidence intervals is necessary to determine the real difference because large samples eliminate the test statistic."

    Correct

    As n→∞n \to \infty, SE=σn→0\text{SE} = \frac{\sigma}{\sqrt{n}} \to 0, which forces z(n)→∞z(n) \to \infty and the p-value→0p\text{-value} \to 0 for any non-zero difference Δ\Delta. Thus, any trivially small difference inevitably becomes statistically significant, necessitating effect sizes and confidence intervals to evaluate practical importance.

    Because of the earlier reciprocal slip, your interpretation was inverted. Large sample sizes do not eliminate the test statistic; they shrink the standard error to zero, making even a negligible effect statistically significant.

WHAT YOU DID WELL
Compliments from Your Teacher

You correctly identified the definition and formula for Cohen's d=Δσd = \frac{\Delta}{\sigma}.

You recognized that Cohen's dd is a population parameter independent of sample size nn.

You correctly related the upper-tail pp-value to the cumulative standard normal distribution.

THE ERROR
Where Your ACEpath Teacher Spotted the Mistake
Complex fraction simplification error

You placed n\sqrt{n} in the denominator instead of the numerator when simplifying Δσ/n\frac{\Delta}{\sigma / \sqrt{n}}.

WHY IT HAPPENED
ACEpath Teacher's Diagnosis

Dividing by a fraction abc\frac{a}{\frac{b}{c}} is equivalent to multiplying by its reciprocal, a⋅cba \cdot \frac{c}{b}. Forgetting to invert the denominator fraction caused the test statistic to decay with sample size rather than scale with n\sqrt{n}.

REASONING PATH
Your ACEpath teacher’s suggested approach
Simplifying Standard Errors in Test Statistics
  • DifferenceSE=Δσn=Δnσ\frac{\text{Difference}}{\text{SE}} = \frac{\Delta}{\frac{\sigma}{\sqrt{n}}} = \frac{\Delta \sqrt{n}}{\sigma}

  • Always rewrite division by a fraction as multiplication by its reciprocal: AB/C=A⋅CB\frac{A}{B/C} = A \cdot \frac{C}{B}.

Large-Sample Behavior of $p$-values
  • As sample size nn increases, standard error shrinks toward 0, causing the test statistic for any real difference Δ≠0\Delta \ne 0 to grow toward ±∞\pm\infty and p→0p \to 0.

  • Sanity check asymptotic behavior: larger samples always provide more power to detect differences, never less.

CORRECTED PATH
ACEpath Teacher's Demonstration
  1. Simplify the test statistic correctly by inverting the denominator: z(n)=Δσn=0.051.00n=0.05nz(n) = \frac{\Delta}{\frac{\sigma}{\sqrt{n}}} = \frac{0.05}{\frac{1.00}{\sqrt{n}}} = 0.05\sqrt{n}.

  2. Evaluate the limits: lim⁡n→∞z(n)=lim⁡n→∞(0.05n)=∞\lim_{n \to \infty} z(n) = \lim_{n \to \infty} (0.05\sqrt{n}) = \infty, which yields lim⁡n→∞p-value=lim⁡z→∞[1−Φ(z)]=0\lim_{n \to \infty} p\text{-value} = \lim_{z \to \infty} [1 - \Phi(z)] = 0.

  3. Explain the core concept: because z(n)→∞z(n) \to \infty, any non-zero difference Δ>0\Delta > 0 will achieve p<αp < \alpha as n→∞n \to \infty. Therefore, statistical significance alone cannot tell you if an effect is meaningful; you must report effect sizes (like d=0.05d = 0.05) and confidence intervals to judge practical significance.

Final answer

(a) z(n)=0.05nz(n) = 0.05\sqrt{n}, d=0.050d = 0.050; (b) lim⁡z(n)=∞\lim z(n) = \infty, lim⁡p=0\lim p = 0, lim⁡d=0.050\lim d = 0.050; (c) As n→∞n \to \infty, SE→0\text{SE} \to 0, forcing z→∞z \to \infty and p→0p \to 0, which guarantees statistical significance for any Δ≠0\Delta \ne 0 regardless of practical importance.

NEXT STEPS
Final Takeaways from Your Teacher
  • Review fraction division rules: ab/c=a⋅cb\frac{a}{b/c} = \frac{a \cdot c}{b}.

  • Practice sketching how the sampling distribution of Xˉ\bar{X} narrows as nn grows.

  • Contrast statistical power (1−β→11-\beta \to 1 as n→∞n \to \infty) with effect size (constant regardless of nn).

Lock it in

Similar question

01

A researcher tests H0:μ=50H_0: \mu = 50 versus Ha:μ>50H_a: \mu > 50 with known σ=10\sigma = 10. If the true mean is μ=51\mu = 51, calculate: (1) Cohen's dd, (2) the test statistic zz when n=100n = 100, and (3) the test statistic zz when n=10,000n = 10{,}000.

Answer

(1) d=μ−μ0σ=110=0.10d = \frac{\mu - \mu_0}{\sigma} = \frac{1}{10} = 0.10. (2) For n=100n = 100, SE=10100=1\text{SE} = \frac{10}{\sqrt{100}} = 1, so z=51−501=1.00z = \frac{51 - 50}{1} = 1.00 (p≈0.1587p \approx 0.1587, not significant at α=0.05\alpha = 0.05). (3) For n=10,000n = 10{,}000, SE=1010,000=0.1\text{SE} = \frac{10}{\sqrt{10{,}000}} = 0.1, so z=51−500.1=10.00z = \frac{51 - 50}{0.1} = 10.00 (p<10−20p < 10^{-20}, highly significant).

02

In an educational study with n=250,000n = 250{,}000 students, an online intervention raises exam scores by Δ=0.2\Delta = 0.2 points on a 100100-point scale with known σ=15.0\sigma = 15.0 points. (a) Compute zz and state whether the result is statistically significant at α=0.01\alpha = 0.01 (zcrit=2.326z_{\text{crit}} = 2.326). (b) Compute Cohen's dd and evaluate whether the intervention has practical significance.

Answer

(a) SE=15.0250,000=15.0500=0.030\text{SE} = \frac{15.0}{\sqrt{250{,}000}} = \frac{15.0}{500} = 0.030. Thus z=0.20.030≈6.67>2.326z = \frac{0.2}{0.030} \approx 6.67 > 2.326. The result is highly statistically significant (p<10−10p < 10^{-10}). (b) d=0.215.0≈0.013d = \frac{0.2}{15.0} \approx 0.013. Because d=0.013d = 0.013 is far below the conventional threshold for a small effect (d=0.20d = 0.20) and represents only 0.2%0.2\% of a standard deviation, the intervention lacks practical significance despite its overwhelming statistical significance.

03

Two studies evaluate the same drug (H0:μ=0H_0: \mu = 0 vs. Ha:μ>0H_a: \mu > 0, σ=4\sigma = 4). Study A has n=16n = 16 and observes xˉ=2.0\bar{x} = 2.0. Study B has n=10,000n = 10{,}000 and observes xˉ=0.1\bar{x} = 0.1. For each study, find the 95%95\% confidence interval for μ\mu and the test statistic zz. Explain which study shows a clinically meaningful effect and which shows statistical significance.

Answer

For Study A: SE=416=1.0\text{SE} = \frac{4}{\sqrt{16}} = 1.0. Test statistic z=2.01.0=2.00z = \frac{2.0}{1.0} = 2.00 (p=0.0228p = 0.0228, significant at α=0.05\alpha = 0.05). 95%95\% CI: 2.0±1.96(1.0)=[0.04,3.96]2.0 \pm 1.96(1.0) = [0.04, 3.96]. Effect size d=2.04=0.50d = \frac{2.0}{4} = 0.50 (medium, clinically meaningful), but the estimate has high uncertainty. For Study B: SE=410,000=0.04\text{SE} = \frac{4}{\sqrt{10{,}000}} = 0.04. Test statistic z=0.10.04=2.50z = \frac{0.1}{0.04} = 2.50 (p=0.0062p = 0.0062, highly significant). 95%95\% CI: 0.1±1.96(0.04)=[0.022,0.178]0.1 \pm 1.96(0.04) = [0.022, 0.178]. Effect size d=0.14=0.025d = \frac{0.1}{4} = 0.025 (negligible). Study B is statistically significant but practically trivial, while Study A shows a clinically meaningful effect that warrants replication with a narrower interval.

Loading workspace…