Participants in Dr. Michèle Youngleson’s game ranked themselves in order of height so that the person in the middle represented the median height.

When Dr. Michèle Youngleson, Senior Improvement Advisor for the South Africa Tuberculosis Quality Improvement (SATBQI) initiative and IHI faculty member, found that learners in her training session struggled with the concept of a median – key to their quality improvement (QI) work with run charts – she developed a new game.

In early February, the SATBQI team conducted its first Leading and Facilitating Quality Improvement workshop for the scale up phase of the project in Johannesburg, facilitated by the IHI team with updates from the National Department of Health. The workshop trained 95 participants from 10 districts during a critical time in the project.

As part of the event, Dr. Youngleson asked seven participants to line up and engaged them in a conversation about the difference between means and medians:

• How would we calculate the mean (average) height of the seven people? (The group correctly answered they would add the heights and divide the total by seven.)
• How would we find the median height? (The group suggested the median is the “middle number” and selected the person with three people on either side.)
• What if the participants lined up in a different order? (Dr. Youngleson swapped the middle person with a participant of a different height.)

“Does this mean that the median is not stable?” Dr. Youngleson asked. The group said no; changing the person standing in the middle does not change the median. Then, to find the middle number that does not depend on the way in which participants happen to line up, the group ranked themselves in order of height so the participant with three people on either side represented the median.

Dr. Youngleson shared her reflections on developing this exercise and using it multiple times to help the concept stick:

What challenge of teaching statistics led you to create this idea?

In health care here in South Africa, you are working mostly with public health nurses and management, who generally just use averages (means). People don't commonly use medians. However, because we use the run chart rule for quality improvement, medians will be used, so we have to teach the concept.

How did you come up with this specific exercise?

The teams really enjoy learning through games. Being in a workshop for a couple of days can be quite tedious if it’s very theoretical, but games and visual displays always help engage participants in complex issues.

This time we did the game on the first day of the workshop when we were teaching medians. Then, on the second day, when the teams needed to draw a median on a run chart as part of an exercise, everyone had forgotten how to calculate the median, so we redid the game. The combination of drawing their own median on the paper graph, along with the reminder of that representation from the game, helped them get the concept. Once didn't seem to be enough.

Would you say that participants reacted positively to the exercise? And if so, was there any other feedback you received besides it being enjoyable?

Yes, they did. They told us that they really enjoyed it and that it was meaningful and helpful. Especially when they were making the graph themselves and placing a median.

There are two things to focus on: that it’s different from an average and that there’s an order to find the middle value, that selecting the middle person is not correct unless they are ranked in order (height order, in this case).

The last part of that demonstration is that the data is not going to present to you in an ordered form but randomly, so the number representing the median can be anywhere in that series of data points.

How did this exercise connect to the teaching methods you typically use?

To make this game easy, I always use an uneven number of participants so it's easy to see the median. I usually use 13, because I want to tell them they need 12 points for a baseline (although I used seven this time).

The game is further supported by an exercise with four sets of numbers on a slide for which teams calculate the median and compare their answers. This is a way of introducing more complex things, such as how do we find a median for an even set of numbers. Or, when there are two or more of the same number in the series, what do we do? Do we count them as one data point or do we count them individually?

What advice would you give to health professionals, educators, or advisors looking to implement new creative ideas like you did with this?

You can find playful ways of conveying complex concepts. The other thing to remember is that repetition is important and there are always different ways of learning.

Editor’s note: This interview has been edited for length and clarity.