One-Sample T-Test Calculator

What Is a One-Sample T-Test?

The one-sample t-test determines whether a sample mean differs significantly from a hypothesized population mean.

Question answered: "Is the true mean different from this specific value?"

Example: "Is the average test score different from the passing threshold of 70?"

The Hypotheses

Null hypothesis ($H_0$): $\mu = \mu_0$ (population mean equals hypothesized value)

Alternative hypothesis ($H_1$):

• Two-tailed: $\mu \neq \mu_0$ (mean is different)
• Right-tailed: $\mu > \mu_0$ (mean is greater)
• Left-tailed: $\mu < \mu_0$ (mean is less)

The Test Statistic

$t = \frac{\bar{x} - \mu_0}{s / \sqrt{n}}$

Where:

• $\bar{x}$ = sample mean
• $\mu_0$ = hypothesized population mean
• $s$ = sample standard deviation
• $n$ = sample size

The t-statistic follows a t-distribution with $df = n - 1$.

Example Calculation

Scenario: Testing whether average daily steps differs from the recommended 10,000.

Data: 25 participants, mean = 8,450 steps, SD = 2,100 steps

Calculation:

$t = \frac{8450 - 10000}{2100 / \sqrt{25}} = \frac{-1550}{420} = -3.69$

P-value: For $df = 24$ (two-tailed), $p \approx 0.001$

Conclusion: The sample mean is significantly below 10,000 steps ($p < 0.01$).

Applications

Quality Control

"Does the production process hit the target weight?"

$H_0: \mu = 500g$

Sample: 30 products, mean = 497g, SD = 8g

Test whether mean differs from target.

Educational Assessment

"Do students perform differently from the national average?"

$H_0: \mu = 75$ (national average)

Sample: Class mean = 81, SD = 12, n = 28

Health Research

"Is the average blood pressure in this population different from normal?"

$H_0: \mu = 120$ mmHg

Sample: Mean = 128, SD = 15, n = 45

Process Monitoring

"Has the mean shifted from the established baseline?"

Compare current sample to historical standard.

Interpreting Results

P-Value Interpretation

Confidence Interval Connection

A 95% CI that excludes $\mu_0$ corresponds to $p < 0.05$.

$CI = \bar{x} \pm t^* \times SE$

If $\mu_0$ is outside this interval, reject $H_0$.

One-Tailed vs. Two-Tailed

Two-tailed (default): Tests for difference in either direction. Use when either increase or decrease is interesting.

One-tailed: Tests for difference in a specific direction. Has more power but requires prior justification.

Two-tailed p-value = 2 × one-tailed p-value

Assumptions

Normality

The population should be approximately normally distributed, OR the sample should be large enough ($n > 30$) for the Central Limit Theorem to apply.

Check with:

• Histogram or QQ-plot
• Skewness and kurtosis
• Formal normality tests (Shapiro-Wilk)

Independence

Observations should be independent—one measurement shouldn't influence another.

Random Sampling

Sample should be randomly selected from the population of interest.

When to Use One-Sample T-Test

Appropriate:

• Comparing sample mean to a known standard
• Testing against theoretical value
• Comparing to published norm
• Process control against target

Not appropriate:

• Comparing two sample means (use two-sample t-test)
• Non-numeric data (use chi-square)
• Severely non-normal data with small n (use Wilcoxon)

Effect Size: Cohen's d

Beyond significance, report effect size:

$d = \frac{\bar{x} - \mu_0}{s}$

In the steps example: $d = \frac{8450 - 10000}{2100} = -0.74$ (medium-large effect)

Historical Context

William Gosset (Student) developed the t-test in 1908 at Guinness Brewery.

The one-sample case was his original focus: testing whether batch quality met standards when sample sizes were small.

His work enabled rigorous inference with limited data—previously, large samples were required for valid conclusions.

Reporting Results

APA format:

"A one-sample t-test revealed that mean daily steps (M = 8,450, SD = 2,100) was significantly lower than the recommended 10,000 steps, t(24) = -3.69, p = .001, d = -0.74."

Always include:

• Sample statistics (M, SD, n)
• Test statistic and df
• P-value
• Effect size

Related Calculations

• Two-Sample T-Test — Compare two groups
• Paired T-Test — Repeated measures
• Confidence Interval for Mean — Interval estimate
• T-Distribution Calculator — P-values
• Cohen's d Calculator — Effect size

Summary

The one-sample t-test determines whether a sample mean significantly differs from a hypothesized value. The test statistic divides the difference (sample mean minus hypothesized mean) by the standard error, producing a t-value compared against the t-distribution. Assumptions include normality (or large n), independence, and random sampling. Always report effect sizes alongside p-values to communicate practical significance, and use confidence intervals to show the range of plausible population means.