Z Score of -0.6: Percentile and What It Means

Home » Blog » Z Score of -0.6: Percentile and What It Means

A z score of -0.6 is approximately the 27.4th percentile in a standard normal distribution. In other words, about 27.4% of values fall below it, while about 72.6% fall above it.

What the Percentile Means

A percentile describes a position within a distribution: the 27.4th percentile is a value equal to or greater than roughly 27.4% of observations in the reference group. It does not mean the score is 27.4% of the maximum, nor that the person or result is “27.4% correct.” Percentile is a rank, not a percentage score.

The interpretation depends on the distribution being used. The 27.4% figure applies when z scores are interpreted against the standard normal curve, a bell-shaped distribution with mean 0 and standard deviation 1. If a z score was calculated from a particular dataset, its percentile under that dataset may differ if the data are not normally distributed.

How a Z Score Works

A z score tells how far a value is from a distribution’s mean, measured in standard deviations. The general formula is:

z = (value − mean) ÷ standard deviation

A z score of -0.6 therefore means the value is six-tenths of one standard deviation below the mean. It is below average, but not unusually far from average in a normal distribution. The negative sign indicates direction; it does not mean the underlying value itself is negative.

For example, suppose a normally distributed test has a mean of 100 and a standard deviation of 15. A score with z = -0.6 corresponds to:

100 + (-0.6 × 15) = 91

That score of 91 is approximately at the 27.4th percentile if the normal model is appropriate. The same z score can correspond to a different raw score in another setting because the mean and standard deviation may differ.

How to Find the Percentile

The percentile associated with a z score is the cumulative probability to its left under the standard normal curve. In notation, it is Φ(z), where Φ is the standard normal cumulative distribution function. For z = -0.6:

  • Look up -0.60 in a standard normal table, or use a calculator or statistics software.
  • Read the cumulative probability: Φ(-0.60) ≈ 0.2743.
  • Convert the probability to a percentile: 0.2743 × 100 ≈ 27.43.

Depending on the table and rounding convention, you may see 27.4%, 27.43%, or 27th percentile. These are the same result at different levels of precision. A z table that reports area between the mean and the z score may instead show about 0.2257; because -0.6 is below the mean, subtract that area from 0.5 to get 0.2743.

Reading the Result in Context

Percentile rank can be useful when comparing a result with a reference group, but it should be interpreted with care. If the reference distribution is normal, a z score of -0.6 places the result below the midpoint yet still within the broad central range of the curve. It is not equivalent to being 60% below average: the z score measures standard deviations, while the percentile measures the proportion below a point.

Nor does a percentile alone show whether a result is important. In education, health, quality control, or another field, the meaning of a score depends on what was measured, how reliable the measurement is, and which comparison group was used. A percentile is descriptive; it does not by itself diagnose a problem or explain why a result occurred.

Two related distinctions help avoid misreading the number:

  • Below the mean is not the same as rare. A z score of -0.6 is moderately below the mean, not far into the lower tail.
  • Percentile rank is not a percent correct. A person at the 27.4th percentile could have answered many or few questions correctly, depending on the test and its scoring.
  • The percentile is not symmetric as a number. The corresponding z score of +0.6 is about the 72.6th percentile, not 27.4. The two locations mirror each other around the mean.

Comparisons With Nearby Z Scores

Nearby z scores show how quickly percentile rank changes around the center of a normal distribution. Approximate cumulative percentiles include:

  • z = -1.0: 15.9th percentile
  • z = -0.6: 27.4th percentile
  • z = 0: 50th percentile
  • z = +0.6: 72.6th percentile
  • z = +1.0: 84.1st percentile

The gaps between these percentiles are not uniform for equal changes in z. The normal curve is densest around its mean, so a small movement near zero can correspond to a larger change in percentile rank than the same movement far out in a tail. Percentiles describe accumulated area, not equal-sized measurement steps.

When the Normal Assumption Matters

The standard normal lookup gives the exact theoretical percentile only when the distribution is normal, or when a method legitimately models the score using a normal distribution. A z score can still be calculated for skewed or irregular data, but converting it through a normal table then gives a model-based estimate rather than necessarily the observed rank.

If you have the original data and want the empirical percentile, compare the value with the observations directly. The precise convention for ties and percentile calculation can vary between software and statistical methods. For a small sample, the empirical rank may also be coarse: one observation can noticeably change the estimated percentile.

Frequently Asked Questions

Is a z score of -0.6 below average?
Yes. It is 0.6 standard deviations below the mean. Under a normal distribution, approximately 27.4% of values are lower and 72.6% are higher.

Is -0.6 the 27th or 28th percentile?
Its cumulative percentile is about 27.4. It is commonly reported as the 27th percentile, or rounded to the 27th or 28th depending on the reporting convention. The unrounded probability is approximately 0.2743.

What percentile is a z score of +0.6?
It is approximately the 72.6th percentile. The standard normal curve is symmetric, so the area below +0.6 is the same as the area above -0.6.

Does a z score of -0.6 mean a score of 40%?
No. It means the value is 0.6 standard deviations below the mean. The corresponding percentile is about 27.4 under a normal model, and neither figure is a percentage-correct score.

Can I use this percentile for any dataset?
Use it as a normal-curve estimate only when the normal assumption is reasonable. For a non-normal dataset, calculate the value’s rank from the actual observations or use a method suited to that distribution.

Related

Scroll to Top