1 Definition

1.1 Basic meaning

A decile is a value that divides an ordered data set into ten equal parts. After the observations are arranged from lowest to highest, each segment contains about 10% of the data. The first decile marks the point below which 10% of observations lie, the second below which 20% lie, and so forth through the ninth decile, which corresponds to 90%.

Deciles are used to describe position within a distribution rather than to measure the average or the total. They help show how values are spread across the full range of a dataset.

1.2 Deciles as quantiles

Deciles belong to the broader family of quantiles, which are values that split ordered data into equal-sized parts. In this framework, deciles divide the distribution into ten sections, making them a specific type of quantile. They provide a compact way to describe relative standing and are often used when a full list of observations would be unwieldy.

1.3 Relationship to percentiles

Percentiles divide data into one hundred equal parts, while deciles divide it into ten. Each decile corresponds to a group of ten percentiles. For example, the third decile aligns with the 30th percentile. Because percentiles offer finer detail, deciles are often viewed as a simpler summary of the same ranking idea.

1.4 Relationship to quartiles and quintiles

Quartiles divide data into four parts, and quintiles divide it into five. Deciles provide a more detailed partition than quartiles or quintiles but a less granular one than percentiles. These measures are all related ways of expressing the position of a value within an ordered distribution.

2 Calculation

2.1 Ordered data sets

To calculate deciles, the data must first be ordered from smallest to largest. The decile points are then identified by locating the positions that correspond to 10%, 20%, 30%, and so on through 90% of the sample. The exact method used may vary slightly depending on the statistical convention or software package.

2.2 Ungrouped data

2.2.1 Position formulas

For ungrouped observations, deciles are often found using a positional formula based on the sample size. A common approach identifies the rank position of the k-th decile as a fraction of the ordered list. If the desired position falls exactly on an observed value, that value is taken as the decile point.

2.2.2 Interpolation methods

When a decile position falls between two observed values, interpolation may be used. This method estimates the decile by taking a weighted average of the surrounding data points. Interpolation produces a smoother estimate, especially in smaller datasets where discrete jumps between observations can be large.

2.3 Grouped data

2.3.1 Cumulative frequency approach

For grouped data, deciles are estimated from cumulative frequencies. The cumulative counts are examined to find the class interval containing the desired decile position. This approach is useful when only summarized frequency tables are available rather than raw data.

2.3.2 Class interval estimation

Once the relevant class is identified, the decile can be estimated using the class boundaries, the cumulative frequency before the class, and the frequency within the class. This method assumes values are distributed reasonably evenly within the interval, allowing an approximate decile value to be calculated.

2.4 Software and computational methods

Statistical software commonly computes deciles automatically from raw or grouped data. Different programs may use different conventions for ranking and interpolation, so results can vary slightly between systems. For this reason, analysts often note the computational method when reporting deciles.

3 Interpretation

3.1 Decile ranks

A decile rank indicates the relative position of a value in a distribution. Someone in the eighth decile, for instance, is above about 80% of the group. Decile ranks are useful for expressing standing in a way that is easy to compare across individuals or categories.

3.2 Percent of observations below a decile

Each decile point represents a cumulative proportion of the dataset. The first decile leaves 10% of values below it, the fifth decile leaves 50% below it, and the ninth decile leaves 90% below it. This cumulative interpretation makes deciles useful for summarizing how values accumulate across a distribution.

3.3 Comparing values across distributions

Deciles are often used to compare the position of values in different datasets, even when the datasets have different units or scales. For example, a score in the top decile of one test can be compared with a score in the top decile of another, since both indicate a similar relative standing. This makes deciles helpful in standardized reporting and comparative analysis.

4 Applications

4.1 Descriptive statistics

In descriptive statistics, deciles provide a concise summary of distribution shape and spread. They can reveal whether values are clustered near the center, spread evenly, or concentrated at one end. Deciles are especially useful when the mean and median alone do not sufficiently describe the data.

4.2 Educational testing

Deciles are used in educational contexts to describe student performance relative to a reference group. A test score may be reported by decile rank to show whether a student performed in the lower, middle, or upper part of the distribution. This approach is often easier to interpret than raw scores alone.

4.3 Income and wage analysis

In economic statistics, deciles are frequently applied to income and wage data. They help summarize how earnings are distributed across a population by showing values at regular intervals. Analysts may compare lower and upper deciles to describe the degree of dispersion in pay or household income.

4.4 Market research and survey analysis

Deciles are also used in market research and survey work to categorize respondents by spending, engagement, satisfaction, or other measured variables. They allow researchers to identify groups at different levels of activity or preference and to compare patterns across segments of a sample.

5 Decile ranges and summaries

5.1 Interdecile range

The interdecile range is the difference between the ninth decile and the first decile. It measures the spread of the middle 80% of the data, excluding the lowest 10% and highest 10%. Because it is less sensitive to extreme values than the full range, it can give a clearer sense of typical dispersion.

5.2 Decile tables

A decile table lists the values corresponding to each decile of a distribution. Such tables provide a compact summary of how observations are distributed across the scale. They are often used in reports that need a straightforward description of ranking or concentration.

5.3 Decile plots

Decile plots display the decile points visually, allowing easy comparison of intervals between them. These plots can highlight asymmetry, clustering, or unusually wide gaps in the data. They are particularly useful when the goal is to compare distributions across multiple groups.

6 Advantages and limitations

6.1 Strengths in data summarization

Deciles are useful because they provide more detail than quartiles while remaining simpler than percentiles. They can summarize a distribution without requiring every observation to be listed. Their rank-based nature also makes them easy to interpret in comparative settings.

6.2 Sensitivity to sample size

In small samples, deciles may be unstable because each observation has a relatively large effect on the result. A small change in the data can shift the estimated decile noticeably. Larger samples usually produce more reliable decile estimates.

6.3 Effects of skewed distributions

In skewed distributions, the distances between deciles may be uneven. Several deciles may cluster closely together in one region and spread widely in another, reflecting the underlying shape of the data. This can be informative, but it also means deciles should be interpreted with attention to distributional form.

7.1 Quantiles

Quantiles are values that divide ordered data into equal-sized groups. Deciles are one common type of quantile and form part of the same conceptual framework.

7.2 Percentiles

Percentiles divide data into one hundred parts and provide a finer ranking scale than deciles. They are often used in score reporting, rankings, and statistical summaries.

7.3 Median

The median is the middle value of an ordered dataset and corresponds to the fifth decile. It divides the distribution into two equal halves and is a key measure of central tendency.

7.4 Histogram and cumulative frequency curve

A histogram shows the frequency distribution of values across intervals, while a cumulative frequency curve shows how counts accumulate as values increase. Both can be used to identify or estimate deciles and to visualize the structure of a distribution.