MCAT topics → Research Methods and Biostatistics → Measures of Spread (Variance, SD, Range, IQR)

Measures of Spread (Variance, SD, Range, IQR)

MCAT trap: Treats variance and standard deviation as equivalent measures with the same units. Variance is expressed in squared units of the data, while standard deviation is the square root of variance and shares the same units as the data.

Measures of spread are tested on the MCAT in passage context — and the most common error is defaulting to SD for everything. SD uses every data point in its calculation, which means one extreme outlier dramatically inflates it, making it misleading for skewed distributions. When data is skewed or contains outliers, IQR is the correct choice — it only considers the middle 50% of data, so extreme values have zero influence on it. The MCAT will present a skewed distribution and ask which spread measure a researcher should report; choosing SD when IQR is appropriate is a reliable wrong-answer trap. The four main measures are range, IQR, variance, and SD — each with a specific appropriate use case.

The trickiest part isn't the math — it's the logic. Students routinely confuse variance and SD as interchangeable, not realizing variance lives in squared units (e.g., kg²) while SD is back in the original units (kg). Similarly, many students default to SD for everything, not recognizing that SD is only appropriate when data is roughly symmetric and normally distributed. Skewed distributions or datasets with outliers demand IQR, which is resistant to extreme values. The MCAT rewards students who can look at a distribution and immediately know which spread measure belongs.

Range is the simplest measure — max minus min — but it's also the most fragile. One outlier completely distorts it. Boxplots are a favorite visual format on the exam: you need to read off the median, IQR (the box itself, from Q1 to Q3), and spot outliers as isolated dots beyond the whiskers. Comparing spread across two boxplots is a classic passage question. If you can interpret a boxplot cold and explain why a researcher chose IQR over SD for their dataset, you're in good shape.

Common misconceptions

Common mistake
Wrong: Variance and standard deviation are interchangeable measures expressed in the same units as the data.
Right: Variance is expressed in squared units of the data, while standard deviation is the square root of variance and shares the same units as the data.
Variance is the average of squared deviations from the mean, so its units are always the square of the original measurement — if you're measuring weight in kg, variance is in kg². SD is simply the square root of variance, which brings the units back to kg. This matters practically: variance is not interpretable in the same scale as your data, which is exactly why SD is reported in research. They measure the same underlying spread, but they are not interchangeable numerically or in terms of units.
Common mistake
Wrong: Standard deviation is the preferred measure of spread for skewed data because it accounts for all values.
Right: IQR is preferred for skewed data or data with outliers because it is resistant to extreme values, while SD is appropriate for symmetric, normally distributed data.
SD uses every data point in its calculation, which sounds like a strength — but it means one extreme outlier can dramatically inflate it, making it misleading for skewed distributions. IQR only cares about the middle 50% of the data (Q1 to Q3), so outliers in the tails have zero influence on it. When a passage describes income, reaction times, or any biologically skewed variable, IQR is the honest measure of spread; SD would overstate variability driven by a few extreme values.
Common mistake
Wrong: The range is a robust measure of spread because it captures the full extent of the data.
Right: The range is the least robust measure of spread because it depends entirely on the two most extreme values and is highly sensitive to outliers.
The range sounds comprehensive because it spans the entire dataset, but that's exactly the problem — it is defined by only two values, the minimum and maximum, both of which are the most likely to be outliers. One erroneous data entry or one biological extreme completely changes the range while leaving the rest of the distribution untouched. A robust statistic resists the influence of outliers; by that definition, range is the least robust spread measure, while IQR is among the most robust.
Free Deck audit

See if your Anki deck covers this topic.

Upload your deck →
Guided session

Stuck on this? An AI tutor that probes your understanding.

Start a session →

What the exam tests

  1. Know the definition and units of each spread measure: range (max − min, same units as data), IQR (Q3 − Q1, same units), variance (average squared deviation, units are squared), and SD (square root of variance, same units as data).
  2. Calculate variance and SD from a small dataset by hand, and understand why the denominator is n−1 (sample) versus n (population) — the MCAT won't always give you a formula, so knowing the logic matters.
  3. Read a boxplot to extract the median, Q1, Q3, IQR, and identify outliers; compare the spread of two distributions using their boxplot features.
  4. Select the correct spread measure for a given research context: use IQR when data is skewed or contains outliers, use SD when data is symmetric and normally distributed.

Can you avoid these mistakes?

A researcher reports that income data from a clinical trial is heavily right-skewed. Should they report SD or IQR as their measure of spread, and why?
A dataset has values: 2, 4, 4, 6, 8, 10. Calculate the mean, then compute the sample variance (using n−1). What are the units of variance versus SD if the values are in centimeters?
You're looking at two boxplots side by side. Box A spans from Q1=20 to Q3=50. Box B spans from Q1=30 to Q3=40. Which group has greater spread, and what is each group's IQR?
A dataset has a range of 200 but an IQR of 12. What does this pattern tell you about the distribution, and which measure better represents the typical spread of the data?

Related topics

Measures of Central Tendency (Mean, Median, Mode) Study Designs (Cross-Sectional, Case-Control, Cohort, RCT) Sampling Methods (Random, Stratified, Cluster, Convenience) Variables (Independent, Dependent, Control, Confounding)

See how your Anki deck covers this topic.

Upload your deck for a free audit →