If .98 means a cumulative probability of 0.98 in the standard normal distribution, its z score is approximately 2.05 (more precisely, about 2.054). In other words, about 98% of values lie below a point 2.05 standard deviations above the mean. If .98 is instead a raw data value, its z score cannot be determined without the distribution’s mean and standard deviation.
What “z score for .98” can mean
The question has two common interpretations. In a z-table or inverse-normal calculation, .98 is usually a probability, and the task is to find the standard normal value whose cumulative area to the left is .98. That value is about 2.05. In a data-analysis problem, .98 might be an observation; converting it to a z score requires the mean and standard deviation of the data.
It is important to identify which quantity .98 represents. A probability is an area under a distribution curve, between 0 and 1. A z score is a location measured in standard deviations. A raw observation, such as a test score of 0.98, has units and does not by itself reveal how unusual it is.
How to find the z score for a cumulative probability of .98
For a standard normal random variable Z, the cumulative distribution function Φ(z) gives the probability that Z is less than or equal to z. We want the value satisfying Φ(z) = 0.98. Looking this up in a cumulative z-table or using an inverse-normal function gives z ≈ 2.0537, commonly rounded to 2.05.
- Confirm that .98 is the area to the left, or cumulative probability.
- Find .9800 in a cumulative standard normal table, then read the corresponding row and column.
- Alternatively, enter invNorm(0.98), or use Φ−1(0.98), in a suitable calculator or statistics program.
- Round as required by the problem, often to 2.05 or 2.054.
Some tables report the area between the mean and z rather than the cumulative area to the left. For .98 to the left, the area from the mean to the z score is .98 − .50 = .48. In that table format, look for approximately .4800; it also leads to z ≈ 2.05.
What the result means
The standard normal distribution has mean 0 and standard deviation 1. A z score of about 2.05 is therefore about 2.05 standard deviations above the mean. The area to its left is 0.98, while the area to its right is about 0.02, or 2%.
This is a percentile interpretation: the point is at approximately the 98th percentile of the standard normal distribution. It does not mean that the point is 98% of the way through the distribution, nor does it mean that 98% of observations equal or fall near that value. It describes the proportion below the point under the assumed model.
For a real normal variable X with mean μ and standard deviation σ, the corresponding value is x = μ + zσ. For example, if test scores are modeled as normal with mean 70 and standard deviation 10, the 98th-percentile score is approximately 70 + (2.054 × 10) = 90.54. This result depends on the normal-model assumption; an actual dataset’s percentile may differ.
Other meanings of .98 and common mistakes
Not every z-score question uses left-tail cumulative probability. The same number can refer to a right-tail area, a central area, or a raw observation. Those interpretations produce different answers.
- .98 to the left: z ≈ 2.054.
- .98 to the right: the left-tail area is .02, so z ≈ −2.054.
- .98 in the middle, centered around zero: each tail has area .01, giving bounds of approximately −2.326 and +2.326.
- .98 as a raw observation: use z = (x − μ) / σ; the mean and standard deviation are needed.
A frequent error is to read .98 as the area between zero and z. That would mean a total left-tail area of 1.48, which is impossible. If a table gives area from the mean, the relevant entry for a left-tail probability of .98 is .48, not .98. Another common slip is to report a positive value for a right-tail probability of .98; a large area to the right requires a negative z score.
Percentiles, cutoffs, and confidence levels
The z value 2.054 is useful for upper-tail cutoffs and percentile calculations, but it should not be confused with the familiar 1.96 cutoff. A central 95% interval under the standard normal curve extends from about −1.96 to +1.96. By contrast, the 98th percentile is a one-sided cutoff: 98% lies below it and 2% above it.
A central 98% interval leaves 1% in each tail, so its endpoints are approximately −2.326 and +2.326. Thus, “98%” alone does not specify a unique z score for an interval: you must know whether it means a one-sided cumulative percentile or a central coverage level. In statistical inference, the appropriate cutoff also depends on the test design, assumptions, and whether the analysis is one-sided or two-sided.
Frequently Asked Questions
Is the z score for .98 exactly 2.05?
Not exactly. The inverse standard normal value is about 2.0537. It is commonly rounded to 2.05, or to 2.054 when three decimal places are requested.
What is the z score if .98 is a right-tail probability?
If 0.98 lies to the right, only 0.02 lies to the left. The corresponding z score is about −2.054.
How do I find the z score for a raw value of 0.98?
Use z = (x − μ) / σ, substituting x = 0.98 along with the distribution’s mean μ and standard deviation σ. Without those values, there is no single answer.
What percentile is z = 2.05?
Its cumulative percentile is about 98%, meaning roughly 98% of a standard normal distribution lies below it. The remaining 2% lies above it.
Is a z score of 2.05 unusual?
Under a standard normal model, a value at or above 2.05 occurs about 2% of the time in the upper tail. Whether that counts as unusual depends on the context and the decision rule; it does not, by itself, establish a cause or prove a result is important.
Quick answer and takeaway
For a cumulative left-tail probability of .98, the standard normal z score is approximately 2.054, usually rounded to 2.05. Before using it, check whether .98 is a left-tail probability, a right-tail probability, a central coverage level, or a raw observation—the interpretation determines the answer.
