Contact
12 programs
Explore the site

Zoho Sheet TDIST: One-Tailed and Two-Tailed Student t Probabilities

Statistical Intermediate Zoho Sheet

Calculate one- or two-tailed Student t probabilities in Zoho Sheet, with validated examples, tail selection and common errors.

Verification status: Official documentation referenced; example outcomes independently reasoned or arithmetically checked where stated. Not executed in Zoho Sheet. How formula examples are checked

Quick answer. =TDIST(2;10;2) computes the two-tailed probability beyond a t statistic of magnitude 2 with 10 degrees of freedom. Its independently calculated expectation is 0.073388035. The corresponding one-tailed formula, =TDIST(2;10;1), has expectation 0.036694017. The difference is the probability in one positive tail versus both symmetric tails; it does not indicate two different definitions of a p-value. Zoho documents these argument positions and mode values [Z-TDIST]. Numerical values here were calculated outside Zoho Sheet [M1].

Official syntax and the three required arguments

TDIST(x; r; mode)
Position Parameter Zoho-documented purpose Safe input guidance
1 x Nonnegative t statistic at which to evaluate the distribution Use ABS(t_statistic) for a two-tailed test; one-tailed tests also require the pre-specified direction, as explained below. Zoho documents #NUM! for x < 0.
2 r Degrees of freedom Use a positive integer obtained from the statistical design; Zoho documents #NUM! when r < 1.
3 mode Tail selection: 1 one-tailed, 2 two-tailed Required. Other values are documented to produce #NUM!.

Zoho’s public Help syntax does not put brackets around any of the three arguments: do not assume mode can be omitted or silently defaults to two tails. Semicolons follow the official English-language examples; entry punctuation and decimal formatting may depend on worksheet locale. No account-plan or first-supported-version date is stated on the function article. These are documented claims, not tests across every platform, browser or regional configuration [Z-TDIST].

Reproducible worksheet: t probability lookup

Enter the following numbers into a new worksheet. Column D contains the corresponding formula, while column E gives a mathematical expected value for a full-precision statistical t distribution.

Row A: absolute t (x) B: df (r) C: mode D: formula E: independently expected result
2 2 10 1 =TDIST(A2;B2;C2) 0.036694017385
3 2 10 2 =TDIST(A3;B3;C3) 0.073388034771
4 2.5 9 2 =TDIST(A4;B4;C4) 0.033861827683
5 0 5 1 =TDIST(A5;B5;C5) 0.500000000000
6 0 5 2 =TDIST(A6;B6;C6) 1.000000000000

The expected outputs follow the standard t survival function: one tail P(T > x) and two tails 2 × P(T > x) for x ≥ 0, with T a Student-t variable of r degrees of freedom. They are model calculations, not observed results from the Zoho web application [M1]. A difference in the last few decimal places is possible because of displayed precision or implementation details; a material difference requires checking parameters and modes.

Important: A two-tailed result is based on the magnitude of the observed statistic. It is not the probability that the population has exactly the null-hypothesis mean, and it does not measure the probability that a hypothesis is true.

Realistic worked example: compare a sample mean with a benchmark

A fictional team records these ten values in A2:A11: 46, 49, 53, 50, 51, 54, 48, 52, 55, 52. The benchmark mean is 50. Assume independent observations and approximate normality of the underlying distribution, with the population standard deviation unknown.

Cell What to enter Independently calculated expectation
B2 =COUNT(A2:A11) 10
B3 =AVERAGE(A2:A11) 51
B4 =STDEV(A2:A11) 2.7888667551
B5 =(B3-50)/(B4/SQRT(B2)) 1.1338934190
B6 =TDIST(ABS(B5);B2-1;2) 0.2861381142
B7 =TDIST(ABS(B5);B2-1;1) 0.1430690571

The sample size is 10, so the one-sample mean test has 10−1=9 degrees of freedom. The two-tailed question asks whether the mean differs from 50 in either direction; its expected p-value is approximately 0.2861. At a conventional 5% decision threshold the evidence in this fictional example is not sufficient to reject a mean of 50. The one-tail calculation would be relevant only if the directional hypothesis had been chosen in advance and its direction matched the positive observed statistic. Do not choose the smaller tail after looking at the data to manufacture significance. The t statistic, standard deviation and p-values were independently computed [M1].

For a negative B5, the use of ABS still yields the two-tailed p-value. For a specified lower-tail directional hypothesis with a negative t statistic, the symmetry permits use of TDIST(ABS(B5);df;1) for the lower tail; selecting the opposite direction would instead require the complement of that one-tail probability. Record the direction before analyzing the sample.

Advanced uses that explain the probability, not just the button

1. Recover a probability from an absolute t threshold

Place 2.228138851986 in F2 and 10 in F3. =TDIST(F2;F3;2) has a model expectation near 0.05, showing why a two-tailed critical value at that threshold corresponds to a 5% test. If a small rounding difference remains, use a more precise value or calculate a threshold with the companion TINV(0.05;10) function [M1; Z-TINV].

2. Compute one- and two-sided reports for transparency

For a positive t statistic and a pre-specified upper-tail alternative, one-sided p is TDIST(x;df;1); the two-sided p is TDIST(x;df;2). For the same positive x they should differ by a factor of approximately two in the mathematical model. A result of 0.04 in one tail is not automatically a two-tail result of 0.04. For x=2 and df=10, compare 0.036694 and 0.073388, not two rounded figures from unrelated rows [M1].

3. Combine with other legitimate test statistics

A one-sample t statistic uses (mean−hypothesized_mean)/(sample_sd/SQRT(n)) and df n−1. A paired-difference analysis first subtracts each matched pair, then applies the one-sample calculation to the differences. A two-sample Welch t-test uses a different standard error and fractional effective degrees of freedom; Zoho’s documentation does not say whether TDIST truncates fractional df, so never assume precise Welch results can be reproduced by simply pasting fractional degrees of freedom. For built-in paired and unequal-variance testing consult Zoho TTEST, whose parameters are different [Z-TTEST].

Errors, limits, and troubleshooting

Symptom What the official reference establishes Action
#NUM! x < 0, r < 1, or mode not 1 or 2 Use an absolute t magnitude; recompute positive degrees of freedom; use exactly mode 1 or 2.
#VALUE! Incorrect argument data type Replace text such as "two" with the numeric mode 2; verify inputs are numbers.
#NAME! Function or named expression not recognized Check spelling TDIST, separators and quotation marks; do not replace it with another platform’s dotted name without Zoho evidence.
#REF! Invalid cell/range reference Repair deleted or moved references.
#N/A! Generic lookup-not-found row in Zoho’s common error table The Zoho article gives no TDIST-specific reproduction for this status. Check data and do not ascribe a particular numerical cause without a native test.
Implausibly large p Wrong tail or degrees of freedom Check mode, correct df and that B5 holds the t statistic rather than the raw measurement.
One-sided probability seems halved Using mode=1 rather than 2 Report the directional hypothesis explicitly and use the matching tail.

Zoho documents the error classes in its Help article, but does not define rounding/truncation of fractional degrees of freedom, behavior at enormous values, or exact numeric precision for every cloud tenant. Those remain verification questions. Blank-cell/coercion behavior should not be inferred from Microsoft Excel, Google Sheets, or mathematical libraries [Z-TDIST].

Compatibility and environment

Zoho’s article documents TDIST(x; r; mode) with modes 1 and 2. Microsoft’s legacy Excel TDIST(x,deg_freedom,tails) is described as a compatibility function, while modern Excel documents T.DIST.RT for one upper tail and T.DIST.2T for two. Google Sheets’ published function list also includes TDIST, but its example notation uses commas. Matching terminology is not a Zoho native execution test, and use of dotted Microsoft names directly in Zoho has not been validated [Z-TDIST; MS-TDIST; MS-TDIST2T; GS-LIST].

The reference reviewed is Zoho’s English-language public web Help article as accessible 2026-10-13. It does not label an earliest supported release, guarantee parity with mobile/offline clients, document a paid account requirement, or certify all locales. There was no logged-in Zoho test session. Validate in the target account and locale before using a result in compliance, financial, or high-consequence statistical reporting.

Review checklist

Before reuse: verify independence and model assumptions; determine one- vs two-tailed hypothesis before seeing results; calculate proper degrees of freedom; use a nonnegative observed t magnitude; compare a row of your own expected values; and label displayed results by probability convention. If a result changes after a Zoho release, capture the actual app build/account/locale, formula and output, then document the official-source comparison rather than overwriting earlier validation records.

Checked references (claim-level evidence)