Descriptive vs Inferential Statistics, Explained with Real Data

Home › Research Methodology › Descriptive vs Inferential Statistics, Explained with Real Data

Quick answerDescriptive statistics summarise the data you actually have: averages, ranges, percentages, charts. Inferential statistics use that data to draw conclusions beyond it: testing whether a difference is real, or estimating values for a wider population. Description tells you what your sample looks like; inference tells you what you can conclude from it.

Below, we explain both in plain language, using real citation data from ScholarsColab.

Descriptive vs inferential statistics at a glance

Descriptive statistics Inferential statistics
Purpose Summarise and describe the data Draw conclusions and test ideas
Question What does my data look like? Is this difference real? Does it apply more widely?
Common tools Mean, median, mode, standard deviation, percentages, charts t-tests, Mann-Whitney U, chi-square, regression, confidence intervals, p-values
Result Numbers and charts about your sample Statements about a population, with a level of uncertainty

Example: describing, then testing, ScholarsColab’s data

To explain both kinds of statistics, we work through ScholarsColab’s own data: 41,792 papers, mostly published between 1998 and 2026, by 967 ScholarsColab mentors (data updated 15 October 2026).

Step 1
Describe the data: how many times are papers cited? Here are the main descriptive statistics for our mentors’ papers:

StatisticValueWhat it means
Number of papers (n)22,045How many values we summarised
Mean (average)34.8Total citations divided by the number of papers
Median7The middle value: half the papers have fewer citations
Mode0The most common value
Standard deviation306.1How spread out the values are
Lowest and highest0 and 19,177The full range
90th percentile6190% of papers have this many citations or fewer
Papers never cited20.8%Share of papers with zero citations

Step 2
Read the description carefully. For 22,045 papers published 2010–2022, the mean is 34.8 citations but the median is only 7, and the most common value (the mode) is 0. The mean is pulled up by a few very famous papers, while most papers have far fewer citations. When data is lopsided like this (statisticians call it skewed), the median describes a typical paper better than the mean.

Step 3
Ask a question description cannot answer. Review articles have a higher median than original papers:

Median citations by type of paperOriginal research620,930 papersReview articles13980 papers
Median citations per paper, by type of paper. Papers published 2010–2022 by ScholarsColab mentors; type identified from the title. Source: ScholarsColab Research Methods Index.

But is that difference real, or could it be luck? Description cannot tell us. We need inference.

Step 4
Use an inferential test. Because citations are skewed, we use the Mann-Whitney U test, which compares the ranks of values rather than their averages. A Mann-Whitney U test comparing 980 review articles with 20,930 original research papers (published 2010–2022) gives a p-value of less than 0.001. That is below 0.05, so the difference is statistically significant: it is very unlikely to be due to chance alone.

Put simply: if you pick one review and one original paper at random, the review has more citations about 59.3% of the time.
Descriptivewhat the data shows
→
Inferentialwhat we can conclude

Common descriptive statistics

Type Measures Use for
Central tendency Mean, median, mode The typical value
Spread Range, standard deviation, interquartile range How much values vary
Frequencies Counts, percentages Categories, such as gender or paper type
Shape Skewness, histograms Whether data is symmetric or lopsided

Common inferential tests

Question Typical test
Do two groups differ (normal data)? Independent t-test
Do two groups differ (skewed data)? Mann-Whitney U test
Do three or more groups differ? ANOVA (or Kruskal-Wallis for skewed data)
Are two categories related? Chi-square test
How does one variable predict another? Regression

Common mistakes reviewers point out

  • Reporting only means for skewed data, such as income or citations, without medians.
  • Drawing conclusions about a population from descriptive statistics alone.
  • Choosing a test that does not fit the data, such as a t-test on heavily skewed values.
  • Reporting p-values without the size of the difference (the effect size).

Frequently asked questions

What is the difference between descriptive and inferential statistics?

Descriptive statistics summarise the data you have, using measures such as the mean and charts. Inferential statistics use sample data to test hypotheses or draw conclusions about a larger population.

Is the mean descriptive or inferential?

The mean of your sample is a descriptive statistic. Using it to estimate the mean of a whole population, with a confidence interval, is inferential.

When should I use the median instead of the mean?

Use the median when data is skewed or has extreme values, such as citations or income, because a few very large values pull the mean upwards.

Do I need inferential statistics for my thesis?

If you compare groups, test hypotheses or want to generalise beyond your sample, yes. Purely descriptive studies, such as some surveys, may only need descriptive statistics.

What is a p-value?

A p-value is the probability of seeing a difference at least as large as the one observed if there were really no difference. Small p-values, usually below 0.05, suggest the difference is unlikely to be due to chance alone.

Use this data in your own work

Use this data in your own research. Download the aggregated figures behind these charts (counts, medians and shares by type of paper, period, method and discipline) as a CSV file. Please cite it as:ScholarsColab (2026). ScholarsColab Research Methods Index [Data set]. Vidhilekha Soft Solutions Pvt Ltd. Retrieved 15 October 2026, from https://scholarscolab.com/research-methodology/. Licensed under CC BY 4.0.Download the data (CSV) →

Need help applying this to your research?

Discuss your research problem with a PhD mentor, hand the whole project to a research project manager, or post your project and let matched experts apply.

Hire a Research Project Manager Hire Experts Browse Mentor Profiles

Mentors whose published work uses this approach:

More research methodology guides

About the data on this page. The ScholarsColab Research Methods Index is built from the Google Scholar publication lists of PhD researchers who mentor on ScholarsColab, and is updated daily as mentors and publications are added. Methods, designs and paper types are identified from titles, so papers whose titles do not mention them are not counted in those groups. The mentors are not a random sample of all researchers (engineering and computer science are strongly represented), so the figures describe this group of researchers and may not apply to research as a whole.