Skip to content
Celeus

Descriptive summary (mean, SD, SE, quartiles)

Shown in the app as descriptive_summary. Part of the method reference, under Descriptive.

When to use

Summarise one or more chosen NUMERIC columns as numbers: n, missing, mean, standard deviation, variance, standard error of the mean with a t-based confidence interval, median, quartiles (R quantile type 7), IQR and range - optionally split by one grouping variable. Each variable uses its own non-missing values. No hypothesis is tested.

What you provide

  • The numeric variables to summarise: one or more numeric columns.
  • Group (optional): one categorical column.

Options

Missing data

  • Complete case (listwise deletion)

Multiple comparisons

  • Not adjustable for this method. Where it makes post-hoc comparisons, the procedure applies its own correction.

Confidence level

Every interval in the result is reported at the level you choose:

  • 90%
  • 95%
  • 99%

Other methods under Descriptive:

How to read the output

The descriptive summary describes one or more numeric columns you choose, as numbers rather than formatted text: for each column it reports the number of values (n), the number missing, the mean, the standard deviation (SD), the variance, the standard error of the mean (SE), a confidence interval for the mean with its margin of error and the t critical value behind it, the median, the first and third quartiles (Q1, Q3), the interquartile range (IQR), and the minimum and maximum. You can split the summary by one grouping column; each column then gets an Overall row followed by one row per group. No hypothesis is tested, so there is no p-value.

SD and SE answer different questions. The SD describes how spread out individual values are. The SE describes how precisely the mean is estimated, and it shrinks as the sample grows (it is the SD divided by the square root of n). The confidence interval of the mean is built from the SE: the mean plus or minus the t critical value for n - 1 degrees of freedom times the SE. The level is 95% by default and can be set to 90% or 99%. Results show SD and SE in separate, named columns rather than as “mean +/- something”, because a bare plus-or-minus is often read the wrong way.

Quartiles use R’s quantile type 7, which is R’s default and is named in the output and in the Methods sentence. There are several standard ways to compute a sample quartile, and other software can give slightly different quartiles on the same data, especially in small samples.

Missing values are handled per column. Each column is summarised on its own non-missing values, so one column’s numbers never depend on which other columns you selected, and two columns with different missing values can report a different n. A row whose group value is missing still counts in the Overall row. With a single value there is no spread, so the SD, SE and interval are shown as “-”; a column whose values are all equal has an SD of 0.

When to use the descriptive table instead. For categorical columns, or for a formatted Table 1 that chooses a mean (SD) or a median [Q1, Q3] per variable and reports counts with confidence intervals for proportions, use the descriptive table (Table 1). A 0/1 column can be summarised here (its mean is a proportion), but the descriptive table is where a proportion gets its proper interval.

Sources: Hyndman RJ, Fan Y (1996), Sample quantiles in statistical packages, The American Statistician 50(4):361-365, doi:10.1080/00031305.1996.10473566.

Worked examples

Celeus's result on a published or illustrative dataset, next to an independent computation and, where one exists, a published value. The data themselves are not reproduced here.

the integers 1 to 10

  • Source: Illustrative or synthetic example data
  • Licence: synthetic (no third-party content)
  • Reference: Hand computation on the integers 1 to 10; not published data
QuantityCeleusIndependent computationPublished valueAgrees
Number of non-missing values, summary row 11010-Yes
Mean, summary row 15.55.5-Yes
Median, summary row 15.55.5-Yes
First quartile (Q1), summary row 13.253.25-Yes
Third quartile (Q3), summary row 17.757.75-Yes
Minimum, summary row 111-Yes
Maximum, summary row 11010-Yes
Standard deviation (SD), summary row 13.027653.02765-Yes
Standard error of the mean (SE), summary row 10.9574270.957427-Yes
Lower confidence limit of the mean, summary row 13.334153.33415-Yes
Upper confidence limit of the mean, summary row 17.665857.66585-Yes

datasets::morley

  • Source: Published dataset distributed with an R package
  • Licence: base R (GPL)
  • Reference: Michelson (1882), Astronomical Papers 1:109-145
QuantityCeleusIndependent computationPublished valueAgrees
Number of non-missing values, summary row 1100100-Yes
Mean, summary row 1852.4852.4-Yes
Median, summary row 1850850-Yes
First quartile (Q1), summary row 1807.5807.5-Yes
Third quartile (Q3), summary row 1892.5892.5-Yes
Minimum, summary row 1620620-Yes
Maximum, summary row 110701070-Yes
Standard deviation (SD), summary row 179.010579.0105-Yes
Standard error of the mean (SE), summary row 17.901057.90105-Yes
Lower confidence limit of the mean, summary row 1836.723836.723-Yes
Upper confidence limit of the mean, summary row 1868.077868.077-Yes

datasets::morley Speed on the NIST Michelso scale (km/s)

  • Source: Published dataset distributed with an R package
  • Licence: base R (GPL)
  • Reference: Michelson (1882), Astronomical Papers 1:109-145. NIST Statistical Reference Datasets (Standard Reference Database 140), DOI 10.18434/T43G6C
QuantityCeleusIndependent computationPublished valueAgrees
Number of non-missing values, summary row 1100100-Yes
Mean, summary row 1299.852299.852299.852Yes
Median, summary row 1299.85299.85-Yes
First quartile (Q1), summary row 1299.808299.808-Yes
Third quartile (Q3), summary row 1299.892299.892-Yes
Minimum, summary row 1299.62299.62-Yes
Maximum, summary row 1300.07300.07-Yes
Standard deviation (SD), summary row 10.07901050.07901050.0790105Yes
Standard error of the mean (SE), summary row 10.007901050.00790105-Yes
Lower confidence limit of the mean, summary row 1299.837299.837-Yes
Upper confidence limit of the mean, summary row 1299.868299.868-Yes

datasets::morley

  • Source: Published dataset distributed with an R package
  • Licence: base R (GPL)
  • Reference: Michelson (1882), Astronomical Papers 1:109-145
QuantityCeleusIndependent computationPublished valueAgrees
Number of non-missing values, summary row 1100100-Yes
Mean, summary row 1852.4852.4-Yes
Median, summary row 1850850-Yes
First quartile (Q1), summary row 1807.5807.5-Yes
Third quartile (Q3), summary row 1892.5892.5-Yes
Minimum, summary row 1620620-Yes
Maximum, summary row 110701070-Yes
Standard deviation (SD), summary row 179.010579.0105-Yes
Standard error of the mean (SE), summary row 17.901057.90105-Yes
Lower confidence limit of the mean, summary row 1839.281839.281-Yes
Upper confidence limit of the mean, summary row 1865.519865.519-Yes

datasets::morley

  • Source: Published dataset distributed with an R package
  • Licence: base R (GPL)
  • Reference: Michelson (1882), Astronomical Papers 1:109-145
QuantityCeleusIndependent computationPublished valueAgrees
Number of non-missing values, summary row 1100100-Yes
Mean, summary row 1852.4852.4-Yes
Median, summary row 1850850-Yes
First quartile (Q1), summary row 1807.5807.5-Yes
Third quartile (Q3), summary row 1892.5892.5-Yes
Minimum, summary row 1620620-Yes
Maximum, summary row 110701070-Yes
Standard deviation (SD), summary row 179.010579.0105-Yes
Standard error of the mean (SE), summary row 17.901057.90105-Yes
Lower confidence limit of the mean, summary row 1831.649831.649-Yes
Upper confidence limit of the mean, summary row 1873.151873.151-Yes

datasets::ToothGrowth

  • Source: Published dataset distributed with an R package
  • Licence: base R (GPL)
  • Reference: Crampton (1947), J. Nutrition 33(5)
QuantityCeleusIndependent computationPublished valueAgrees
Number of non-missing values, summary row 16060-Yes
Mean, summary row 118.813318.8133-Yes
Median, summary row 119.2519.25-Yes
First quartile (Q1), summary row 113.07513.075-Yes
Third quartile (Q3), summary row 125.27525.275-Yes
Minimum, summary row 14.24.2-Yes
Maximum, summary row 133.933.9-Yes
Standard deviation (SD), summary row 17.649327.64932-Yes
Standard error of the mean (SE), summary row 10.9875220.987522-Yes
Lower confidence limit of the mean, summary row 116.837316.8373-Yes
Upper confidence limit of the mean, summary row 120.789420.7894-Yes
Number of non-missing values, summary row 23030-Yes
Mean, summary row 220.663320.6633-Yes
Median, summary row 222.722.7-Yes
First quartile (Q1), summary row 215.52515.525-Yes
Third quartile (Q3), summary row 225.72525.725-Yes
Minimum, summary row 28.28.2-Yes
Maximum, summary row 230.930.9-Yes
Standard deviation (SD), summary row 26.605566.60556-Yes
Standard error of the mean (SE), summary row 21.2061.206-Yes
Lower confidence limit of the mean, summary row 218.196818.1968-Yes
Upper confidence limit of the mean, summary row 223.129923.1299-Yes
Number of non-missing values, summary row 33030-Yes
Mean, summary row 316.963316.9633-Yes
Median, summary row 316.516.5-Yes
First quartile (Q1), summary row 311.211.2-Yes
Third quartile (Q3), summary row 323.123.1-Yes
Minimum, summary row 34.24.2-Yes
Maximum, summary row 333.933.9-Yes
Standard deviation (SD), summary row 38.266038.26603-Yes
Standard error of the mean (SE), summary row 31.509161.50916-Yes
Lower confidence limit of the mean, summary row 313.876713.8767-Yes
Upper confidence limit of the mean, summary row 320.049920.0499-Yes

datasets::airquality

  • Source: Published dataset distributed with an R package
  • Licence: base R (GPL)
  • Reference: Chambers, Cleveland, Kleiner and Tukey (1983), Graphical Methods for Data Analysis
QuantityCeleusIndependent computationPublished valueAgrees
Number of non-missing values, summary row 1116116-Yes
Mean, summary row 142.129342.1293-Yes
Median, summary row 131.531.5-Yes
First quartile (Q1), summary row 11818-Yes
Third quartile (Q3), summary row 163.2563.25-Yes
Minimum, summary row 111-Yes
Maximum, summary row 1168168-Yes
Standard deviation (SD), summary row 132.987932.9879-Yes
Standard error of the mean (SE), summary row 13.062853.06285-Yes
Lower confidence limit of the mean, summary row 136.062436.0624-Yes
Upper confidence limit of the mean, summary row 148.196248.1962-Yes
Number of non-missing values, summary row 22626-Yes
Mean, summary row 223.615423.6154-Yes
Median, summary row 21818-Yes
First quartile (Q1), summary row 21111-Yes
Third quartile (Q3), summary row 231.531.5-Yes
Minimum, summary row 211-Yes
Maximum, summary row 2115115-Yes
Standard deviation (SD), summary row 222.224422.2244-Yes
Standard error of the mean (SE), summary row 24.358574.35857-Yes
Lower confidence limit of the mean, summary row 214.638714.6387-Yes
Upper confidence limit of the mean, summary row 232.59232.592-Yes
Number of non-missing values, summary row 399-Yes
Mean, summary row 329.444429.4444-Yes
Median, summary row 32323-Yes
First quartile (Q1), summary row 32020-Yes
Third quartile (Q3), summary row 33737-Yes
Minimum, summary row 31212-Yes
Maximum, summary row 37171-Yes
Standard deviation (SD), summary row 318.207918.2079-Yes
Standard error of the mean (SE), summary row 36.06936.0693-Yes
Lower confidence limit of the mean, summary row 315.448615.4486-Yes
Upper confidence limit of the mean, summary row 343.440343.4403-Yes
Number of non-missing values, summary row 42626-Yes
Mean, summary row 459.115459.1154-Yes
Median, summary row 46060-Yes
First quartile (Q1), summary row 436.2536.25-Yes
Third quartile (Q3), summary row 479.7579.75-Yes
Minimum, summary row 477-Yes
Maximum, summary row 4135135-Yes
Standard deviation (SD), summary row 431.635831.6358-Yes
Standard error of the mean (SE), summary row 46.20436.2043-Yes
Lower confidence limit of the mean, summary row 446.337446.3374-Yes
Upper confidence limit of the mean, summary row 471.893471.8934-Yes
Number of non-missing values, summary row 52626-Yes
Mean, summary row 559.961559.9615-Yes
Median, summary row 55252-Yes
First quartile (Q1), summary row 528.7528.75-Yes
Third quartile (Q3), summary row 582.582.5-Yes
Minimum, summary row 599-Yes
Maximum, summary row 5168168-Yes
Standard deviation (SD), summary row 539.681239.6812-Yes
Standard error of the mean (SE), summary row 57.782137.78213-Yes
Lower confidence limit of the mean, summary row 543.93443.934-Yes
Upper confidence limit of the mean, summary row 575.989175.9891-Yes
Number of non-missing values, summary row 62929-Yes
Mean, summary row 631.448331.4483-Yes
Median, summary row 62323-Yes
First quartile (Q1), summary row 61616-Yes
Third quartile (Q3), summary row 63636-Yes
Minimum, summary row 677-Yes
Maximum, summary row 69696-Yes
Standard deviation (SD), summary row 624.141824.1418-Yes
Standard error of the mean (SE), summary row 64.483024.48302-Yes
Lower confidence limit of the mean, summary row 622.265222.2652-Yes
Upper confidence limit of the mean, summary row 640.631340.6313-Yes
Number of non-missing values, summary row 7146146-Yes
Mean, summary row 7185.932185.932-Yes
Median, summary row 7205205-Yes
First quartile (Q1), summary row 7115.75115.75-Yes
Third quartile (Q3), summary row 7258.75258.75-Yes
Minimum, summary row 777-Yes
Maximum, summary row 7334334-Yes
Standard deviation (SD), summary row 790.058490.0584-Yes
Standard error of the mean (SE), summary row 77.453297.45329-Yes
Lower confidence limit of the mean, summary row 7171.2171.2-Yes
Upper confidence limit of the mean, summary row 7200.663200.663-Yes
Number of non-missing values, summary row 82727-Yes
Mean, summary row 8181.296181.296-Yes
Median, summary row 8194194-Yes
First quartile (Q1), summary row 87272-Yes
Third quartile (Q3), summary row 8284.5284.5-Yes
Minimum, summary row 888-Yes
Maximum, summary row 8334334-Yes
Standard deviation (SD), summary row 8115.075115.075-Yes
Standard error of the mean (SE), summary row 822.146322.1463-Yes
Lower confidence limit of the mean, summary row 8135.774135.774-Yes
Upper confidence limit of the mean, summary row 8226.819226.819-Yes
Number of non-missing values, summary row 93030-Yes
Mean, summary row 9190.167190.167-Yes
Median, summary row 9188.5188.5-Yes
First quartile (Q1), summary row 9127127-Yes
Third quartile (Q3), summary row 9270.75270.75-Yes
Minimum, summary row 93131-Yes
Maximum, summary row 9332332-Yes
Standard deviation (SD), summary row 992.88392.883-Yes
Standard error of the mean (SE), summary row 916.95816.958-Yes
Lower confidence limit of the mean, summary row 9155.484155.484-Yes
Upper confidence limit of the mean, summary row 9224.85224.85-Yes
Number of non-missing values, summary row 103131-Yes
Mean, summary row 10216.484216.484-Yes
Median, summary row 10253253-Yes
First quartile (Q1), summary row 10175175-Yes
Third quartile (Q3), summary row 10273273-Yes
Minimum, summary row 1077-Yes
Maximum, summary row 10314314-Yes
Standard deviation (SD), summary row 1080.568380.5683-Yes
Standard error of the mean (SE), summary row 1014.470514.4705-Yes
Lower confidence limit of the mean, summary row 10186.931186.931-Yes
Upper confidence limit of the mean, summary row 10246.037246.037-Yes
Number of non-missing values, summary row 112828-Yes
Mean, summary row 11171.857171.857-Yes
Median, summary row 11197.5197.5-Yes
First quartile (Q1), summary row 11107107-Yes
Third quartile (Q3), summary row 11231231-Yes
Minimum, summary row 112424-Yes
Maximum, summary row 11273273-Yes
Standard deviation (SD), summary row 1176.834976.8349-Yes
Standard error of the mean (SE), summary row 1114.520414.5204-Yes
Lower confidence limit of the mean, summary row 11142.064142.064-Yes
Upper confidence limit of the mean, summary row 11201.651201.651-Yes
Number of non-missing values, summary row 123030-Yes
Mean, summary row 12167.433167.433-Yes
Median, summary row 12192192-Yes
First quartile (Q1), summary row 12116.75116.75-Yes
Third quartile (Q3), summary row 12234.5234.5-Yes
Minimum, summary row 121414-Yes
Maximum, summary row 12259259-Yes
Standard deviation (SD), summary row 1279.118379.1183-Yes
Standard error of the mean (SE), summary row 1214.44514.445-Yes
Lower confidence limit of the mean, summary row 12137.89137.89-Yes
Upper confidence limit of the mean, summary row 12196.977196.977-Yes

a single value

  • Source: Illustrative or synthetic example data
  • Licence: synthetic (no third-party content)
  • Reference: Hand computation on a one-row frame; not published data
QuantityCeleusIndependent computationPublished valueAgrees
Number of non-missing values, summary row 111-Yes
Mean, summary row 177-Yes
Median, summary row 177-Yes
First quartile (Q1), summary row 177-Yes
Third quartile (Q3), summary row 177-Yes
Minimum, summary row 177-Yes
Maximum, summary row 177-Yes

Validation report

References

  1. Hyndman, R. J. and Fan, Y. (1996). Sample quantiles in statistical packages. The American Statistician 50: 361-365. doi:10.1080/00031305.1996.10473566 doi:10.1080/00031305.1996.10473566