How PSI Works

The principles underlying the Personal Sanity Index are drawn from the laws of statistics and their Quality Assurance applications in manufacturing.

As an engineer, I used the tools of statistical process control for years. The reason most of the world’s leading companies use these techniques is because they work. And they’re reliable and simple to use.

So, how does it work?

First, you need a process or activity that is regular and repeatable. A machine that makes the same part over and over. Or your daily commute to work. For each of these examples, there’s an average measurement one could take: inches for the part or time for the commute. If you take daily measurements and plot them on a chart, you’d likely see something like the illustration in Figure 1 (I’ll use the example of commuting since it’s something most of us have experienced).

There are a few things we can observe right away. First, no two measurements (in this case, commuting time) are exactly the same. Some days it takes a little more time; other days a little less. That’s normal, natural variation. Just as the daily temperature varies daily in a given season.

The other main feature of this type of graph is that there is an average or typical amount of time for the commute. The same is true of temperature. In July, temperatures are warmer than January, but the monthly average is pretty similar from one year to the next.

It may seem simple and obvious, but those two features – average and daily variance – are common to most things in our lives. And that’s what we’ll see if we build an effective PSI.

We want to include factors that we often accomplish, but acknowledge that we may not achieve them every day. If we include items in our Index that we never achieve (or we always do), then we won’t have any variance.

For example, if you think meditation is a good idea, but you’ve never done it and you don’t even know how to start, that might not be a great item to include. But if meditation is something you often do, but not always, then it could be a great candidate.

That’s the thinking behind the Core Four. We know that we are healthier when we sleep well, eat well, exercise, and connect with others. But we don’t always hit our goals with each of them, or all of them, every day. If we try to improve our consistency with those behaviors, our average should increase, but there will still be daily variance.

And that’s one important way charts can help us – we can see changes in average.

I live in a college town, so there aren’t many people around in the summer. The first chart that showed my commuting time included data for two weeks in the middle of August. But look what happens when we get to the end of summer and students return for fall classes (Figure 2).

More people means more cars on the road, which leads to longer commuting times. There’s still variation from day to day, depending on whether I hit red or green lights, or if there are more or fewer people using crosswalks, or whether it’s raining. But it takes consistently longer in September than it does in August. That is caused by an external influence – in this case, students returning to campus.

The PSI can show us how certain factors influence our overall state of health and well-being. The first year I lived in Ann Arbor, I rented a house that was close to a railway crossing. Sometimes late at night, a train would roll through, sounding its horn and jolting me awake. That negatively effected my sleep and kept my PSI average score low.

Since then, I moved to a quieter part of town, well away from the tracks and my sleep improved dramatically. Some nights, I still don’t get a great sleep – so there’s still variance – but the average is definitely better!

Sometimes shifts in average can indicate a permanent or long-term change; other times there may be a temporary cause to the shift. Figure 2 also shows a week where the average commute time increased by several minutes and then returned to the previous level. The cause, in this case, was road construction.

We can experience temporary disruptions in many aspects of our lives. During my years as a recreational athlete, I’ve injured knees, hips, shoulders, and other body parts. Sometimes I could replace one type of exercise with another.

For example, during rehab for a knee injury, I could often swim. But sometimes that wasn’t possible, and I simply had to wait (not very patiently) for the injury to heal. That would reduce my PSI average (and usually also increase my weight), but once I returned to my normal routines, the PSI would also bounce back to its previous level (though the weight loss often took a little longer).

There are two additional signals that we can pick up from a well-designed chart. One is trends and the other is change in variance.

Trends can alert us to changes as they are starting to happen. In the commuting example, you can see the trend to longer travel times began during the last week of August. As more students returned, traffic volume increased and it took a little longer each day to get to work. Eventually, the trend leveled off at the new normal, and regular variation returned. Trends are typically indicated by three or more consecutive observations in the same direction; increasing or decreasing. You can think of daily variation like tossing a coin. Heads and tails will roughly alternate and the long-term average will be half heads and half tails. It’s unusual to get several heads or tails in a row. Sometimes it’s just random, but it will often indicate something is changing and it’s worth paying attention.

The road construction didn’t signal the shift with a leading trend – it just happened. But it also went away again just as suddenly. Trends aren’t always there to warn us that a change is coming. But if we see a pattern in our PSI where the daily scores are steadily changing, it might be worth taking some time to figure out why.

Increasing workload sometimes sneaks up on us. We do a little more each day to try to stay on top of things, or to catch up when we fall behind. Maybe that means skipping a workout. Or grabbing some fast food on the way home. Or staying up later than usual.

These factors will add up and it’ll be obvious if we are paying attention to the PSI chart. We can’t always do anything to change the demands of work, but sometimes we can. And, at the very least, being aware that we’re not looking after ourselves as well as we should might help us make the extra effort to squeeze in a workout, eat a healthy meal, or get to bed a little earlier.

So far, we’ve been looking at changes in average. Those changes can be short- or long-term, sudden or gradual. And the charts can help us see them when they happen, though it’s up to us to figure out the cause. Sometimes it’s obvious (e.g., road construction) while in some cases we never quite understand what has caused a change to take place.

Change in variance is something else altogether. In many ways, it’s harder to spot and the consequences of increased variance can be severe. Returning one final time (I promise) to the commuting example, you can see in Figure 2 that the range of daily times is much larger in late September than it was earlier that month. It’s become less predictable. Less regular. And that increases the chances of arriving late to work. It’s harder to plan for big swings than it is to work within a regular set of expected outcomes.

In the manufacturing environment, increased variance can be a huge problem. Most processes have tolerances, measured in millimeters or grams or some other metric.

As long as the variation is within tolerance – within the control limits – then you can have confidence that the parts are being made correctly.

But when variance takes us outside those limits, you can’t count on the process any longer and you have to measure every single part to ensure no bad ones are getting through. That can cost a manufacturer a lot of time and money! Losing control can be a nightmare.

The same can be said for our well-being, as measured by the PSI. If I am in a regular set of rhythms and routines with an average score of four that’s up or down by one or two points each day, that’s a pretty good place for me to be. But if the daily ups and increase to plus or minus three points with no predictability, that is the definition of out-of-control. Especially if I don’t know what’s causing it.

Increased variance usually has a cause, just like changes in average have a cause. And sometimes the causes are obvious. For a machine tool, it could be as simple as a loose guide. Once the problem is discovered, the errant part can be fixed and the process brought back under control.

I’ve had times when increased PSI variance can be traced to an obvious cause. Hosting family members in my home for an extended period can really throw a wrench into daily routines, but at least that’s easy to understand and has an end in sight. But when there’s no clear reason for the change, that is perhaps the most critical time to invest in self-discovery to track down the cause(s) and bring things back under control. Friends and family can be great resources at times like this. Things that are invisible to us can be blindingly obvious to the people around us.

Figure 3 captures all of the elements we’ve been discussing in my Personal Sanity Index. I’ve pulled this together from different times in my life (remember, I’ve been doing this for a couple of decades now), but all of the examples are real. If I can see a trend or a shift, I at least have a chance to do something about it. But if it’s invisible to me, I won’t. For me, that’s the real value – and magic – of the PSI.

In my personal journey, I’ve seen disruptions caused by work travel that affect sleep, exercise, nutrition, and my connection to others. It’s easy to spot during the second week in Figure 3. Another disruption can be observed in week four when I the house was full of loving but demanding family members. On the other hand, vacation often provides an opportunity to invest time in self-care (week five) and my daily Index scores are frequently highest when I’m away from the normal demands of work (though I don’t always have the healthiest eating habits when on holiday).

In summary, our well-being is reflected in, and influenced by, our daily behaviors. Keeping an eye on patterns, especially trends and changes in averages, can help us maintain a healthier, happier and more balanced state. It’s been helping me though 30 years of ups and downs and I hope it can also be of some benefit to you.