Z Test

Introduction

Z test: parametric test assessing population mean differences when population variance is known. Based on normal distribution properties. Critical in inferential statistics for validating hypotheses about means or proportions. Utilized extensively in quality control, clinical trials, and social sciences.

"Statistics is the grammar of science." -- Karl Pearson

Definition and Purpose

Definition

Z test: statistical hypothesis test evaluating whether sample mean differs significantly from population mean, assuming known population variance and normality or large sample size (n > 30).

Purpose

Objective: test null hypothesis (H0) about population mean or proportion against alternative (H1). Decide acceptance or rejection of H0 based on z statistic and significance level (α).

Scope

Applicable to: single sample mean test, difference between two means with known variances, and proportion tests.

Assumptions

Known Population Variance

Population standard deviation (σ) must be known and constant.

Normality

Population distribution normal or sample size large enough (Central Limit Theorem) for normal approximation.

Random Sampling

Sample drawn randomly and independently from population.

Scale of Measurement

Data measured at interval or ratio level.

Types of Z Tests

One-Sample Z Test

Tests if sample mean equals hypothesized population mean.

Two-Sample Z Test

Compares means of two independent samples with known variances.

Z Test for Proportions

Tests hypotheses about population proportions using normal approximation to binomial distribution.

Z Test Statistic Formula

One-Sample Z Test Formula

z = (x̄ - μ₀) / (σ / √n)

where x̄ = sample mean, μ₀ = hypothesized population mean, σ = population standard deviation, n = sample size.

Two-Sample Z Test Formula

z = (x̄₁ - x̄₂) / √(σ₁²/n₁ + σ₂²/n₂)

where x̄₁, x̄₂ = sample means, σ₁, σ₂ = population standard deviations, n₁, n₂ = sample sizes.

Z Test for Proportions

z = (p̂ - p₀) / √(p₀(1 - p₀) / n)

where p̂ = sample proportion, p₀ = hypothesized population proportion, n = sample size.

Testing Procedure

Step 1: State Hypotheses

Formulate H0 (null hypothesis) and H1 (alternative hypothesis).

Step 2: Set Significance Level

Choose α (commonly 0.05 or 0.01).

Step 3: Calculate Test Statistic

Compute z using sample data and formulas.

Step 4: Determine Critical Value or P-value

Find z critical values from standard normal table or calculate p-value.

Step 5: Decision

Reject H0 if |z| > z critical or p-value < α; otherwise, fail to reject H0.

Example Calculation

Problem Statement

Sample of n=50 yields mean weight 72 kg. Population mean weight is hypothesized as 70 kg. Population σ=8 kg. Test at α=0.05 if sample mean differs significantly.

Calculation

Compute z:

z = (72 - 70) / (8 / √50) = 2 / (8 / 7.071) = 2 / 1.131 = 1.77

Decision

Critical z for two-tailed α=0.05 is ±1.96. Since 1.77 < 1.96, fail to reject H0.

Interpretation

Insufficient evidence to conclude sample mean differs from population mean at 5% significance level.

Interpretation of Results

Rejecting Null Hypothesis

Indicates data inconsistent with H0; alternative hypothesis supported.

Failing to Reject Null Hypothesis

No strong evidence against H0; cannot confirm difference.

P-value Role

P-value quantifies probability of observing test statistic as extreme assuming H0 true; smaller p-value more evidence against H0.

Confidence Intervals

Z test results align with confidence intervals for population mean.

Advantages and Disadvantages

Advantages

• Simple computation and interpretation.
• Based on well-understood normal distribution.
• Powerful for large samples with known variance.

Disadvantages

• Requires known population variance, often unavailable.
• Less accurate for small samples with unknown variance.
• Assumes normality, which may not hold.

Comparison with Other Tests

Z Test vs. T Test

Z test: known σ, large n; T test: unknown σ, small n; uses Student's t-distribution.

Z Test vs. Chi-Square Test

Z test: mean/proportion comparison; Chi-square: categorical data independence/goodness-of-fit.

Z Test vs. ANOVA

Z test: two groups; ANOVA: more than two groups or factors.

Applications

Quality Control

Monitor process means, detect deviations from target.

Medical Research

Compare treatment effects when variance known or large samples.

Social Sciences

Test population parameters such as means and proportions.

Market Research

Evaluate consumer characteristics against hypothesized values.

Common Misconceptions

Z Test Always Applicable

Incorrect: only valid with known σ or large sample size.

P-value Indicates Probability H0 is True

False: p-value measures data extremeness assuming H0 true, not H0 truth probability.

Failing to Reject H0 Proves H0 True

Incorrect: lack of evidence is not proof of equality.

References

• Wackerly, D. D., Mendenhall, W., Scheaffer, R. L., Mathematical Statistics with Applications, Duxbury Press, 7th ed., 2008, pp. 150-170.
• Devore, J. L., Probability and Statistics for Engineering and the Sciences, Cengage Learning, 9th ed., 2015, pp. 350-375.
• Rice, J. A., Mathematical Statistics and Data Analysis, Duxbury Press, 3rd ed., 2006, pp. 220-245.
• Moore, D. S., McCabe, G. P., Craig, B. A., Introduction to the Practice of Statistics, W. H. Freeman, 8th ed., 2014, pp. 300-320.
• Montgomery, D. C., Introduction to Statistical Quality Control, Wiley, 7th ed., 2012, pp. 95-115.