In the spring of 2003, at the Intel Northwest Science Expo in Portland Oregon, I had the honor of meeting a rather remarkable young fourth grader. She was remarkable because, while most middle and high school students, and arguably most adults, have difficulty internalizing and creating meaning from numerical data, this bright-eyed fourth grader in the white cotton print dress knew exactly what her data was telling her.

When I asked what she had concluded from her experiments, she stood up straight, looked me directly in the eye, and made a sweeping hand gesture to her hand-plotted charts, declaring “Well, as you can clearly see from my data, my hypothesis was completely wrong.”

The Science Project

This young lady’s science project explored the effect of different light-dark cycles on the growth of tomato plants. Her original hypothesis was that plants which have cycles of 2 hours of light followed by 2 hours of dark (the experimental group) would grow slower than those with 12 hours of continuous light followed by 12 hours of continuous dark (the control group).

On first meeting her, I remember being particularly impressed by the fact that she was a fourth-grader competing in a middle school science fair against students two to five years older than her. However, while the work was impressive, a key element in the judging process is determining the amount external help the student received as compared to how much was actually performed by the student, and it was clear that she had received a good deal of help from her father in both the experimental design and the setup.

What made her stand out to the judges that day, what got her into the final round of judging even with the amount of help she received from her father, was the level to which she had assimilated her data. None of the more than 30 middle school students I had spoken with that morning had been so clearly focused on the meaning of their data.

To give an example, many students had utilized three-dimensional bar charts created by Microsoft Excel, but none of the students who used them were able to accurately read those charts. The three-dimensional bars in Excel charts are placed a distance from the background scale, thus creating a parallax error. None of the students corrected for the parallax in reading bar height.

Other students had utilized some of the statical functions provided by excel. But when pressed ever so gently to explain the meaning of the statistics, it was clear that they didn’t really understand what those statics were telling them. One student confided to me, “Well, I don’t really understand the statistics, but Excel did the work for me.”

By contrast, the confident sweeping hand gesture this fourth grader made to her hand-plotted charts, combined with the “as you can clearly see,” suggested that she had fully assimilated and made meaning of her data. It took her all of about 20 seconds to explain her trend lines, and what they were telling us. It turned out that the pattern of light-dark cycles showed no effect on plant growth.

About half way through her time scale however, the growth rate of the experimental group showed a sudden acceleration. When I asked about that, she explained that the point of change signified the end of one experiment, and the beginning of another experiment.

Having concluded from her data that there was no significance to the light-dark cycle timing, she had formulated a new hypothesis, that maybe growth rate was, all else held constant, strictly a function of total hours of light. For her new experiment, she gave the experimental group continuous, 24 hours of light, while the control group continued to receive 12 hours of light followed by 12 hours of dark.

She showed how her data suggested that her new hypothesis was at least partially proven correct. She pointed out however that the slope of the continuous light group was not quite twice that of the control group, suggesting that light was not the only factor involved.

Remember, this was a fourth grader.

Lessons for Educators

Base your work on valid and verifiable data

W. Edward Demming has been quoted as saying something to the effect that you cannot improve what you don’t measure. Of course, there are all kinds of caveats related to how you measure things, not all things can be quantitatively evaluated. But the point remains that to improve, there needs to be some measure, and the measure needs to be consistent.

This fourth grader didn’t get into the best of fair category by saying, “well, I just knew those tomatoes weren’t growing as fast as they should.” Remarkably, I hear things like “I’ll just know when my students are learning” from teachers all the time.

Improvement requires an emotionally safe environment where it is not only OK to admit you are wrong, it is expected

When I was learning to ski, I had an instructor tell me, “if you aren’t falling, you aren’t learning.” In quality assurance circles, tests which always succeed and tests which always fail don’t provide much information. Useful information comes through a balance of success and failure.

There was no shame, no fear of rejection in the bright eyes of this fourth-grader in the white cotton print dress as she stood up, looked me directly in the eye and announced her failure. Rather, she was demonstrating that she existed in an emotionally safe environment, probably both at home and school. As a result of quickly acknowledging her failure, she was able to adapt her experiment to create new knowledge.

But how many teachers would be willing to go into a faculty meeting and say, “well, as you can clearly see from my data, my hypothesis about how to teach this material effectively was completely wrong?”

Let’s take that a step further, how many teachers would invite another teacher or administrator into their classroom to evaluate their technique on a regular basis?

To enable a learning organization, the fear of failure needs to be eliminated. Failure needs to viewed as an opportunity to learn. Learning needs to be a continuous expectation. The only failure to be feared should be failure to learn.

Gaining knowledge from data is like juggling balls, you have to be fully involved

One cannot learn to juggle by reading a book on juggling. You need to hold the balls in your hand, to toss them around. If you practice enough, and drop enough balls, at some point, you will succeed in keeping them suspended in motion. Data is the same way, to really learn from that data, you need to be willing to handle it, to juggle it around.

A school district I worked with had invested a fair amount of money in a comprehensive data management system. I was told that when the data system was fully implemented, they would be able to track student progress and that would result in making adjustments to their teaching.

I have nothing against data systems, they can certainly make life easier for teachers, administrators, students and parents. It is common to see my own kids log into their school grade keeping system to check on assignments, scores and grades. In many ways, it makes life more convenient, it helps in managing the administrivia of school.

However, thinking back to that fourth-grader and her had plotted charts as compared to the competing student’s computer generated charts, I have to wonder if this new data system will really improve student learning. Most of those other middle school students who used computers, couldn’t even read their charts accurately. Those who used Excel’s statical tools could not explain the meaning of those statistics.

I’ve sat in meetings where Excel charts are projected from PowerPoint slides to the nodding of every head in the room. Until I asked the presenter to label the axes of the graph, and they could not, suggesting the graph was meaningless.

Computers may make things too easy, allow one to distance themselves from the data. Often it is better to hand plot data, to create the chart on a white board or butcher paper, to engage other people in the kinesthetic juggling act of meaning making.

Conclusion

Richard Saul Wurman, in his book, Information Anxiety, describes an understanding continuum. Data, as discrete measures of some form sits at one end of the continuum. Data becomes Information when it can be organized into patterns. Information becomes Knowledge when you can make meaning from it. What I learned from a fourth grader that day was the importance of assimilating that data in order to make meaning from it. I focus on Education here, but the lesson is just as important to business and industry. If a fourth grader can do it, we can too.