A z score table tells you the exact probability that a data point falls below, above, or between specific values in a standard normal distribution. If you've ever stared at rows of decimals and wondered how to turn a raw score into a meaningful percentage, you're in the right place.
Put simply, a z score represents how many standard deviations a given value sits away from the mean. When you look up that number in a z score table, you map raw data onto a standard normal curve with a mean of zero and a standard deviation of one. This process lets you compare values from completely different datasets on a single, universal scale.
Understanding the Standard Normal Distribution
Before diving into lookup procedures, you need to understand what the underlying curve represents. A standard normal distribution is symmetrical. It has a total area under the curve equal to one, or one hundred percent. The mean sits dead center at zero.
Because the distribution is symmetrical, exactly fifty percent of the data falls below the mean, and fifty percent falls above it. Standard deviations stretch out symmetrically on both sides. A z score of $1.0$ means your data point is one standard deviation above the mean. A z score of $-1.5$ means your data point is one and a half standard deviations below the mean.
Real-world data rarely arrives pre-standardized. Test scores, heights, financial returns, and manufacturing tolerances all use different units and scales. Standardizing data converts these apples and oranges into comparable units. You can learn more about how standard scores function in theoretical frameworks via the National Institute of Standards and Technology.
How to Calculate a Z Score
You can't use a z score table until you have an actual z score. The math itself is straightforward. You take your raw data point, subtract the population mean, and divide the result by the population standard deviation.
Here is the formula written out:
$$z = \frac{x - \mu}{\sigma}$$
In this equation, $x$ represents your raw score, $\mu$ represents the population mean, and $\sigma$ represents the population standard deviation.
Imagine you take a standardized exam where the average score is $75$ with a standard deviation of $8$. You score an $87$. Let's plug those numbers into the formula. First, subtract $75$ from $87$, which gives you $12$. Next, divide $12$ by the standard deviation of $8$. Your resulting z score is $1.5$. Your score sits one and a half standard deviations above the average.
What happens if your score falls below the mean? Imagine a score of $63$ on that same exam. Subtracting $75$ gives you $-12$. Dividing by $8$ yields a z score of $-1.5$. The negative sign simply tells you the score is below average.
Reading a Z Score Table Correctly
A standard z score table—often called a unit normal table—contains rows and columns of probabilities. Most tables are cumulative from the mean outward, or cumulative from the extreme left tail up to your z score. You need to check which type of table you're holding before you read a single number.
Most standard tables break the z score into two parts. The row header gives you the units and the first decimal place. The column header gives you the second decimal place.
Let's look up our positive z score of $1.5$ from the exam example.
- Find the row labeled $1.5$ (or $1.50$) on the left side of the table.
- Find the column labeled $.00$ at the top of the table.
- Locate the intersection cell.
For a z score of $1.50$, the table value is typically $.9332$. This decimal means that $93.32$ percent of the distribution falls below a z score of $1.5$. Consequently, only about $6.68$ percent of test-takers scored higher than you did.
What about negative z scores? If you're using a cumulative table from the left tail, look up $-1.50$. The table value is $.0668$. This tells you that $6.68$ percent of the population scored lower than a z score of $-1.5$.
Solving Different Probability Scenarios
Real-world questions rarely ask for a simple cumulative probability. You'll often need to find the area between two scores, or the probability that a value exceeds a certain threshold. Here is how you handle those scenarios without getting tripped up.
Finding the Area Above a Positive Z Score
If you want to know the probability of scoring higher than a z score of $1.5$, you can't just take the table value of $.9332$. That number represents everything below the score. Because the total area under the curve equals one, you subtract the table value from one.
$$1 - 0.9332 = 0.0668$$
About $6.68$ percent of the population exceeds this threshold.
Finding the Area Between Two Z Scores
Suppose you want to know what percentage of scores fall between a z score of $-1.0$ and a z score of $2.0$.
- Look up the cumulative probability for the higher z score ($2.0$), which is $.9772$.
- Look up the cumulative probability for the lower z score ($-1.0$), which is $.1587$.
- Subtract the smaller probability from the larger one.
$$0.9772 - 0.1587 = 0.8185$$
About $81.85$ percent of the data falls between these two points. For deep statistical methodologies and dataset guidelines, check resources hosted by the U.S. Census Bureau.
Common Mistakes to Avoid
Even experienced analysts make simple errors when working with normal distributions. Watch out for these common traps.
- Confusing sample standard deviation with population standard deviation. If you're working with a sample rather than an entire population, make sure your formulas and software settings align properly.
- Mixing up left-tail tables and cumulative-from-mean tables. Always check the visual diagram printed at the top of your specific z table.
- Forgetting to convert raw scores. You can't look up a raw score like $87$ directly in a standard table. You must convert it into a z score first.
- Misinterpreting negative signs. A negative z score doesn't mean a negative probability. It just means your data point sits below the mean.
Frequently Asked Questions
Can a z score be higher than 3.0?
Yes. A z score simply measures distance from the mean in standard deviations. While values beyond $3.0$ are rare in a standard normal distribution—occurring less than $0.3$ percent of the time—they happen in skewed or wide datasets.
What is the difference between a z score and a t score?
You use a z score when you know the exact population standard deviation and have a large sample size. You use a t score when the population standard deviation is unknown and you must estimate it using a smaller sample.
How do I convert a z score back into a raw score?
You reverse the original formula. Multiply the z score by the standard deviation, then add the population mean.
$$x = (\mu + (z \times \sigma))$$
Does the underlying data have to be normally distributed?
Using a z score table assumes your data follows an approximately normal distribution. If your dataset is heavily skewed or contains extreme outliers, standard z score lookups will yield inaccurate probability estimates.
