The critical value of z is a threshold on the standard normal distribution that helps researchers decide whether an observed test statistic is extreme enough to reject a null hypothesis. By comparing the calculated z‑score to the critical value, analysts can make objective decisions about statistical significance without relying solely on p‑values. In hypothesis testing, this value separates the region where the null hypothesis is considered plausible from the region where it is deemed unlikely given the chosen significance level (α). Understanding how to determine and interpret the critical value of z is essential for anyone working with confidence intervals, proportion tests, or mean comparisons that assume normality.
Understanding the Standard Normal Distribution
Before diving into critical values, it is useful to recall the properties of the standard normal distribution (often denoted Z ~ N(0,1)). 5th, 5th, 95th, and 97.Which means key percentiles—such as the 2. This distribution has a mean of 0, a standard deviation of 1, and a symmetric bell‑shaped curve. The total area under the curve equals 1, representing the probability of all possible outcomes. 5th—correspond to specific z‑scores that frequently appear in statistical tables.
Because the standard normal distribution is fully defined, any probability statement can be transformed into a z‑score and vice‑versa. This property makes it the backbone of many inferential procedures, especially when sample sizes are large enough for the Central Limit Theorem to apply Most people skip this — try not to..
What Is a Critical Value?
A critical value is a cut‑off point on the sampling distribution of a test statistic. Also, if the observed statistic falls beyond this point (in the direction specified by the alternative hypothesis), the null hypothesis is rejected. And for a z‑test, the critical value is expressed as z₍α/2₎ for two‑tailed tests or z₍α₎ for one‑tailed tests, where α is the pre‑selected significance level (commonly 0. 05, 0.01, or 0.10) Simple as that..
- Two‑tailed test: The rejection region is split between both tails; each tail contains α/2 of the total probability.
- One‑tailed test: The entire α resides in a single tail, either the left or the right, depending on the direction of the alternative hypothesis.
The critical value itself is not a probability; it is a z‑score that marks the boundary where the tail probability equals the designated α (or α/2).
How to Find the Critical Value of Z
Finding the critical value involves locating the z‑score that corresponds to a cumulative probability of 1 − α (for a right‑tailed test) or α/2 (for each tail in a two‑tailed test). The most common tools are:
- Standard normal (z) table – provides the area to the left of a given z‑score.
- Statistical software – functions such as
qnorm(p, mean=0, sd=1)in R orscipy.stats.norm.ppfin Python return the quantile directly. - Calculators – many online calculators allow users to input α and test type to obtain the critical value instantly.
Step‑by‑Step Using a Z‑Table
- Determine α and whether the test is one‑ or two‑tailed.
- For a right‑tailed test, locate the probability 1 − α in the body of the table and read the corresponding z‑score.
- For a left‑tailed test, locate α directly; the resulting z‑score will be negative.
- For a two‑tailed test, compute α/2, find the z‑score that leaves α/2 in the right tail (i.e., cumulative probability 1 − α/2), and then use ± that value as the critical boundaries.
Example Calculations
- α = 0.05, two‑tailed: α/2 = 0.025. Look up 0.975 (1 − 0.025) in the table → z ≈ 1.96. Critical values: −1.96 and +1.96.
- α = 0.01, left‑tailed: Look up 0.01 → z ≈ −2.33. Critical value: −2.33.
- α = 0.10, right‑tailed: Look up 0.90 → z ≈ 1.28. Critical value: +1.28.
One‑Tailed vs Two‑Tailed Tests
Choosing between one‑tailed and two‑tailed tests affects the critical value dramatically because it changes how α is allocated The details matter here..
| Test Type | α Allocation | Critical Value(s) for α = 0.645 |
| Left‑tailed | All α in left tail | −1.05 |
|---|---|---|
| Right‑tailed | All α in right tail | +1.645 |
| Two‑tailed | α/2 in each tail | ±1. |
A one‑tailed test is more powerful for detecting an effect in the predicted direction because the critical value is closer to zero (e.Plus, g. Plus, , 1. 645 vs 1.96). That said, it should only be used when there is a strong theoretical justification for expecting the effect to lie in a single direction; otherwise, a two‑tailed approach is safer.
It sounds simple, but the gap is usually here.
Practical Examples
Example 1: Testing a Population Mean
Suppose a researcher wants to know whether the average score on a standardized exam differs from 75. A sample of n = 100 yields a mean of 78 with a known population standard deviation σ = 10. The test statistic is:
[ z = \frac{\bar{x} - \mu_0}{\sigma/\sqrt{n}} = \frac{78 - 75}{10/\sqrt{100}} = \frac{3}{1} = 3.0 ]
With α = 0.05 and a two‑tailed alternative, the critical values are ±1.96. But since 3. 0 > 1.96, the null hypothesis (μ = 75) is rejected And that's really what it comes down to..
Example 2: Proportion Test – One‑Tail
A company claims that at least 60 % of its customers are satisfied. Because of that, a survey of 200 customers finds 108 satisfied (p̂ = 0. 54). To test whether the true proportion is less than 0 That's the whole idea..
[ z = \frac{\hat{p} - p_0}{\sqrt{p_0(1-p_0)/n}} = \frac{0.54 - 0.Think about it: 60}{\sqrt{0. In real terms, 60·0. 40/200}} = \frac{-0.06}{0.0346} ≈ -1 Worth knowing..
For a left‑tailed test at α =
