From “Higher” to “5 Points Higher”
In Matching Measures: Mean With Standard Deviation, Median With IQR, you learned to use the same appropriate measure of center for both groups. This tutorial takes the next step: quantify how far apart the centers are. Instead of saying one group’s center is “higher,” you can report that it is, for example, 5 points higher, with the units that make the comparison meaningful.
The calculation is subtraction. The interpretation requires a little more care: specify which group is subtracted from which, name the measure of center, and put the difference back into the context of the variable. The difference is measured in the same units as the original observations.
A difference is directional. If you calculate Group B minus Group A and get a positive result, Group B has the higher center by that amount. If you reverse the subtraction, the result has the opposite sign. The summaries have not changed; only the order of the comparison has.
For a clear sentence, translate the arithmetic into words. If the mean for Group B is 81 points and the mean for Group A is 76 points, then \(81-76=5\) points. In context, Group B’s mean score is 5 points higher than Group A’s. The units are points because the original variable is a score measured in points.
A Reliable Way to Calculate and Explain the Difference
Before subtracting, confirm that you are comparing the same variable for the two groups and using the same measure of center. As in the earlier tutorial on choosing summary statistics, inspect the distributions to decide whether means or medians are more suitable. Then use that same type of center for both groups.
This is a descriptive comparison, not an inference procedure. There is no significance test or confidence interval in this calculation. The difference describes the two summaries you have calculated or been given. It does not, by itself, establish why the groups differ or guarantee that every observation in one group is higher than every observation in the other.
Check that both centers describe the same quantitative variable in the same units.
Compare mean with mean or median with median, using the distribution features to guide the choice.
For example, calculate Group B minus Group A if you want to describe how much higher or lower B is than A.
State the difference, direction, variable, groups, and units. Keep the conclusion limited to what the summaries show.
If the difference is negative, that is not a calculation error. It means the first group in your stated order has a lower center than the second group. You can either keep the negative sign and explain it, or reverse the order and report a positive difference with the direction changed accordingly. Do not report a positive difference without making clear which group is higher.
Worked Example: Comparing Mean Quiz Scores
A fictional teacher compares quiz scores, in points, for students in two study groups. The scores are:
Group A: 68, 72, 76, 80, 84 points
Group B: 73, 77, 81, 85, 89 points
Question: How far apart are the groups’ mean quiz scores?
State. We will compare the mean quiz score for Group B with the mean quiz score for Group A. The variable is quiz score, measured in points.
Plan. Both lists are roughly symmetric and have no value far from the rest, so a mean comparison is reasonable. We will calculate each group’s mean, then subtract Group A’s mean from Group B’s mean. This describes the observed scores; it is not a test of whether the groups differ in a larger population.
Do. Group A’s score total is \(68+72+76+80+84=380\). With five students, its mean is \(380/5=76\) points. Group B’s total is \(73+77+81+85+89=405\), so its mean is \(405/5=81\) points. The difference, Group B minus Group A, is
Conclude. In these observations, Group B’s mean quiz score is 5 points higher than Group A’s mean quiz score. This compares the groups’ averages; it does not say that every student in Group B scored 5 points more than a particular student in Group A.
Interpreting Differences Between Medians
The same subtraction idea applies when the appropriate center is the median. As discussed in Matching Measures: Mean With Standard Deviation, Median With IQR, medians are often useful when distributions are skewed or have outliers. Subtract the medians, not one group’s median from the other group’s mean.
When the difference is not a whole number, keep enough precision to represent the result sensibly. A median difference of 2.5 minutes is still measured in minutes. Do not change it into a percent or call it “2.5 times as large”; those are different kinds of comparisons.
Worked Example: Comparing Median Equipment-Inspection Times
A fictional repair workshop records the time, in minutes, needed to inspect equipment at two workstations:
Workstation A: 18, 20, 21, 23, 24, 25, 61 minutes
Workstation B: 16, 18, 19, 20, 22, 24, 28 minutes
Question: Compare the medians, and explain which workstation has the higher center.
Workstation A has an inspection time of 61 minutes, much larger than its other times. That value gives the distribution a long right tail, so the median is a reasonable resistant measure of center. Both lists have seven ordered values, making the fourth value the median.
For Workstation A, the fourth value is 23 minutes. For Workstation B, the fourth value is 20 minutes. Subtracting B’s median from A’s median gives
In these observations, Workstation A’s median inspection time is 3 minutes higher than Workstation B’s median inspection time. The 61-minute observation does not pull the median upward in the way it could influence the mean. The conclusion compares the middle observations; it does not claim that every inspection at A took longer than every inspection at B.
What a Difference in Centers Does and Does Not Tell You
A difference in centers describes the distance between two summaries, not the distance between every pair of observations. For example, two groups can have medians that differ by 4 units even though their individual observations overlap considerably. To describe overlap or spread, you would need to examine other features of the distributions, as in the earlier tutorial on overlap and separation.
A difference also does not identify a cause. If one group’s mean is 5 points higher, the calculation alone cannot show that a particular teaching method, product, or condition caused the difference. State what the summaries show without adding an unsupported explanation.
Finally, a numerical difference needs a useful reference. “The centers differ by 5” is incomplete if the reader does not know the variable or units. “Group B’s mean quiz score is 5 points higher than Group A’s” identifies the measure, direction, amount, and unit. If both groups’ centers are given in seconds, dollars, centimeters, or another unit, retain that unit in the difference.
Worked Example: Reading the Sign When the Order Is Reversed
A fictional community center compares the median wait time for two service desks. Desk Cedar has a median wait of 14 minutes, and Desk Birch has a median wait of 19 minutes. The variable is wait time, measured in minutes.
Question: Calculate Cedar minus Birch and interpret the result. Then express the comparison with the higher center first.
The requested order is Cedar minus Birch:
The negative result means Cedar’s median wait is lower than Birch’s median wait. Specifically, Cedar’s median wait is 5 minutes lower than Birch’s. To state the comparison with the higher center first, reverse the subtraction:
Thus, Desk Birch’s median wait time is 5 minutes higher than Desk Cedar’s. Both statements describe the same difference; the sign and wording change with the subtraction order. The comparison is about median wait times, not a claim that every customer at Birch waited exactly 5 minutes longer.
Common Mistakes and AP Exam Tips
- Leaving the subtraction order unstated. “The difference is 5” does not tell the reader which group has the higher center. Name the order or state clearly which group is higher.
- Dropping the units. A difference of 5 points is not the same as a difference of 5 minutes. Include the units of the original variable.
- Subtracting unlike summaries. Do not compare Group A’s mean directly with Group B’s median. Use means for both groups or medians for both groups, as appropriate.
- Confusing the difference between centers with individual differences. A 5-point difference between means does not mean that every person in one group scored exactly 5 points more than someone in the other group.
- Claiming the difference explains itself. A descriptive difference does not establish what caused it or whether it would persist in a broader population. Keep the conclusion connected to the summaries and the observed groups.
- Reporting only an unsigned amount. If a subtraction gives \(-5\), explain the negative result or reverse the subtraction and adjust the wording. Do not silently turn a negative result into a positive one.
For a full-credit comparison, show which matching centers you subtract, give the arithmetic, and interpret the result with the groups, variable, direction, and units. A concise sentence such as “Group B’s median delivery time is 2.5 minutes lower than Group A’s” communicates the result more clearly than a number without context.
Check Your Understanding
Use the summaries given to calculate and interpret each difference. State the subtraction order in your response.
- Group M has a mean of 42 points and Group N has a mean of 47 points. What is the difference in mean scores, with Group N compared with Group M?
- Two gardens have median plant heights of 18 centimeters and 15.5 centimeters. How much higher is the first garden’s median than the second garden’s median?
- A calculation of Group A minus Group B gives \(-6\) minutes. What does the negative sign tell you about their centers?
- Two groups’ mean temperatures differ by 3 degrees. Does that show that every temperature in one group is 3 degrees higher than a temperature in the other? Explain.
- Why should you not subtract one group’s mean from another group’s median to describe a difference in centers?