A z score of 1.64 means a value is 1.64 standard deviations above the mean of a distribution. In a standard normal distribution, it corresponds to a cumulative probability of about 0.9495: roughly 94.95% of values lie below it and 5.05% lie above it.
What a z score measures
A z score expresses how far a value is from a mean in units of standard deviation. It lets you compare positions within distributions even when their original units or scales differ. A positive score is above the mean; a negative score is below it; zero is exactly at the mean.
For a value x, population mean μ, and population standard deviation σ, the usual formula is:
z = (x − μ) / σ
For example, if test scores have a mean of 70 and a standard deviation of 10, a score of 86 has z = (86 − 70) / 10 = 1.6. That score is 1.6 standard deviations above the mean. The score itself is not 1.6 points above average; the z score measures distance in standard-deviation units.
How to interpret 1.64 and its percentile
When values follow a standard normal distribution—a bell-shaped distribution with mean 0 and standard deviation 1—the z score of 1.64 is near the 95th percentile. More precisely, the standard normal cumulative probability at 1.64 is approximately 0.9495. Thus, about 95 out of 100 observations are below that point and about 5 out of 100 are above it.
“About the 95th percentile” is a useful shorthand, not an exact identity. The z score 1.645 is commonly used as the more precise cutoff for the 95th percentile in a standard normal distribution. Tables and software may round values differently, so 1.64 can appear as 94.95%, 94.9%, or approximately 95%.
The percentile statement depends on the distribution. A z score of 1.64 always means 1.64 standard deviations above the mean, but it only translates to about the 95th percentile if the distribution is normal, or if a suitable normal approximation applies. In a skewed or unusual distribution, the actual proportion below the value may be different.
How to calculate the original value
If you know the mean and standard deviation and want the value associated with z = 1.64, rearrange the formula:
x = μ + zσ
Suppose adult heights in a particular reference group have a mean of 170 cm and a standard deviation of 8 cm. A z score of 1.64 corresponds to 170 + (1.64 × 8) = 183.12 cm. This calculation locates a value relative to those stated statistics; it does not establish that the reference group is normally distributed or that the value is unusual in every context.
For a sample, the corresponding calculation often uses the sample mean and sample standard deviation, written as x̄ and s. Whether to use population or sample statistics depends on the question and the method being applied. In hypothesis tests, a z statistic can also describe a standardized difference, rather than an individual observation.
Why 1.64 appears in statistical tests
The number 1.64 is often encountered as a critical value in one-tailed tests and confidence calculations. For a standard normal distribution, approximately 5% of the area lies above z = 1.645. A one-sided test with a 5% significance level may therefore reject a null hypothesis when its test statistic exceeds about 1.645, provided the test’s assumptions and direction are appropriate.
For a two-sided test with a total 5% significance level, the 5% is split between both tails. The usual standard normal critical values are approximately −1.96 and +1.96, not ±1.64. This distinction matters: using 1.64 as a two-sided 5% cutoff would produce a larger rejection region and a different error rate.
- One-tailed, upper-tail 5% cutoff: about 1.645.
- One-tailed, lower-tail 5% cutoff: about −1.645.
- Two-tailed 5% cutoff: about −1.96 and +1.96.
These are standard normal reference values. Some analyses use a t distribution instead, especially when estimating a mean from a small sample and the population standard deviation is unknown. Its critical value depends on the degrees of freedom and may differ from 1.645.
Using a z table or calculator
A cumulative z table gives the area to the left of a score. Find z = 1.64 in the table to get approximately 0.9495. To find the upper-tail area, subtract that result from 1: 1 − 0.9495 = 0.0505, or about 5.05%.
Some tables show the area between the mean and the z score instead of the cumulative area. For 1.64, that area is about 0.4495; add 0.5 to obtain the area to the left, 0.9495. Check the table’s key before interpreting the entry. A calculator or statistical program can provide the same cumulative probability, often labeled as a normal cumulative distribution function.
Common interpretation mistakes
- Confusing z with a percentage: 1.64 is not 1.64%, nor does it mean 164% of the mean. It is a distance measured in standard deviations.
- Treating the percentile as guaranteed: the 94.95% figure comes from the standard normal model, not from the z score alone.
- Mixing up one and two tails: 1.645 is a familiar one-sided 5% cutoff; a two-sided 5% cutoff is approximately 1.96 in each direction.
- Assuming a high percentile means a large practical difference: the real-world size of the difference depends on the standard deviation and the subject being measured.
A percentile is also not the probability that a particular observation is correct, important, or caused by a specific factor. It describes a position under a chosen distribution or reference population. Interpretation should account for how the data were collected and whether the model is a good fit.
Frequently Asked Questions
Is a z score of 1.64 good?
It means the value is 1.64 standard deviations above the mean. Whether that is good depends on what is measured: a higher score may be favorable for one measure and unfavorable for another.
What percentile is a z score of 1.64?
In a standard normal distribution, it is approximately the 94.95th percentile, often rounded to the 95th percentile. About 5.05% of the distribution lies above it.
What is the difference between z = 1.64 and z = 1.645?
They are close values. A score of 1.645 is a more precise standard normal cutoff for the 95th percentile, while 1.64 gives a cumulative probability of about 94.95%.
Does z = 1.64 mean a 5% significance level?
It is approximately the upper-tail critical value for a one-sided 5% test under a standard normal model. It is not the usual cutoff for a two-sided 5% test, which uses about ±1.96.
How do I find the raw value for z = 1.64?
Use x = μ + 1.64σ, where μ is the mean and σ is the standard deviation. You need both quantities to calculate the raw value.
