# Using Pictures to Communicate

Nothing is easier than misleading with statistics.

The creation of statistics, is about communicating information, but it is the sender’s idea of meaning.. The preliminary part is analyzing data and summarizing it to make some point. Few people understand the analysis part and so often make mistakes deriving meaning from the presentation.

You cannot easily become highly skilled, but it is not hard to know enough to recognize a misleading story. Think about the risk to misunderstanding.

### The kinds of statistics you will see

There are many ways to present.

Common ones are data tables, line graphs, and pie charts. The form relates to the story you are trying to tell. A pie chart is usually about distribution. A line graph usually compares more than one variable and how they compare over time or volume.  A single line will show accumulated growth or rate of change.

People like this sort of presentation because we are conditioned to be visible and these contribute far more information than long descriptions or even data tables. That ease of interpretation is why you need to pay attention. Ask yourself about the meaning intended as well as any other meaning that is possible.

Think through purpose and whether there are other meanings you could apply.

### Graphs are not facts, they are interpretation of facts. That’s different.

Presenting options

Take a line graph of the performance of the S&P500 for the last 90 years. If I want to get you to think the market is all very different now, I would present the graph like this.

S&P 500 90 Year Graph

Courtesy of Macrotrends

This graph displays accumulated value by year. Your automatic response is likely that sometime around 1980, things changed. It became more violent. Must be risky.

There are other ways to present.

Another graph of the same data presents it using a logarithmic scale instead of the linear one above.

Same data, different message. The present is not so different as we thought a few seconds ago. Logarithmic presentation focuses on rate of change, while linear accumulates. This picture tells you that rate of change, ignoring fluctuations, is not so different now. You could draw a line through the fluctuation line and not have many far from it. Rate of change is yield and that is a more important idea for an investor.

If you saw the linear format only you would not likely have an intuitive awareness of the logarithmic one.

Think about this. If the market grows 10% a year, when it is at 4000 it goes up 400 point the next year. That will look big on the linear format, ordinary on the logarithmic one.

The Dow Jones Industrial index fell 2,997 points on March 16, 2020.

It fell 508 points on October 19th, 1987

On October 28th and 29th, 1929 it fell a total of 69 Points.

Which was the worst?

When the market is high it takes large point drops to make it serious. Does the raw number strike fear into your heart? If all you read are headlines, it might.

Think about perspective. Meaning is what you want

First layer of meaning

• 2020 was 2,997 points — 12.93% down.
• 1987 was 508 points  — 22.61% down
• 1929 was  just 69 points, but 23.3% down.

2020 was not such a big deal. If you look, there is a 10% drop about once every 11 months. So not so unusual.

Second layer of meaning

1987 recovered before the government got a chance to “fix” it. 1929 did not.

### The takeaway

Pay attention to the meaning not what the picture tells you at first glance.

Know how statistics fit into your context. If you know there is a 10% drop each year, and you don’t need the money that day, why is it fearsome?

If you expect the market to grow 10% a year before fees and taxes, shouldn’t you pay attention to the logarithmic graph instead of the linear one?

Understand the story being told from the storyteller’s point of view before accepting it as your story.

I help people have more retirement income and larger, more liquid estates.

Call in Canada 705-927-4770, or email don@moneyfyi.com

This site uses Akismet to reduce spam. Learn how your comment data is processed.