Home » Statistics & Probability Calculators » T Value for a 90% Confidence Interval: Critical t by df

T Value for a 90% Confidence Interval: Critical t by df

Home › Statistics & Probability Calculators › t Value for 90% Confidence

t Value for 90% Confidence

For a 90% confidence interval, the critical value t* leaves 0.05 in each tail of the t distribution, so it depends on the degrees of freedom (n − 1 for one sample): t* = T⁻¹(1 − α/2, df), α = 0.1. At df = 10, t* = 1.8125; it shrinks towards the z value, 1.6449, as the sample grows.

Use t instead of z when the standard deviation is estimated from the sample. The table gives two-sided and one-sided values; the calculator gives any df.

Critical t values for 90% confidence

Computed from the inverse t distribution (SciPy).
dfTwo-sided t* (90% CI)One-sided t (90%)
16.31383.0777
22.92001.8856
32.35341.6377
42.13181.5332
52.01501.4759
61.94321.4398
71.89461.4149
81.85951.3968
91.83311.3830
101.81251.3722
111.79591.3634
121.78231.3562
131.77091.3502
141.76131.3450
151.75311.3406
161.74591.3368
171.73961.3334
181.73411.3304
191.72911.3277
201.72471.3253
211.72071.3232
221.71711.3212
231.71391.3195
241.71091.3178
251.70811.3163
261.70561.3150
271.70331.3137
281.70111.3125
291.69911.3114
301.69731.3104
401.68391.3031
501.67591.2987
601.67061.2958
801.66411.2922
1001.66021.2901
1201.65771.2886
∞ (z)1.64491.2816

Worked example: confidence interval for a mean

A class of n = 12 students has a mean score of 74.5 with sample standard deviation s = 8.2.

  1. Degrees of freedomn − 1 = 11
  2. Critical valuet* = 1.7959 (90%, df = 11)
  3. Standard errors ÷ √n = 8.2 ÷ √12 = 2.3671
  4. Margin of error1.7959 × 2.3671 = 4.251
  5. 90% interval74.5 ± 4.251 = (70.25, 78.75)

Other confidence levels

Two-sided critical t values (SciPy).
Confidencedf = 5df = 10df = 20df = 30
80%1.4761.3721.3251.310
90%2.0151.8121.7251.697
95%2.5712.2282.0862.042
98%3.3652.7642.5282.457
99%4.0323.1692.8452.750

Frequently asked questions

What is the t value for a 90% confidence interval?
It depends on the degrees of freedom: 2.015 at df = 5, 1.812 at df = 10, 1.725 at df = 20 and 1.697 at df = 30 (two-sided).
Why is t bigger than z?
Estimating the standard deviation from the sample adds uncertainty, so the t distribution has heavier tails than the normal; the extra width disappears as the sample grows.
What is the one-sided t value at 90% confidence?
1.3722 at df = 10: all of α sits in one tail.

Last reviewed: November 18, 2026. Every figure on this page is calculated from the formulas shown; the calculator runs in your browser and nothing you enter is sent anywhere.

Scroll to Top