Last week, a Charlotte police officer shot and killed a man. The police say he was holding a gun; the man’s family and witnesses say he was holding a book. The body and dash cam footage released as of this weekend (Sunday, 25 Sept) does not provide the necessary angles to see what he had.
But here is the truth of the situation: the officer perceived a gun; the witnesses perceived a book. What we see and what we perceive (that is, how our brains translate what we see) are not the same thing. The brain processes information based on different cognitive biases, and fear and adrenaline further distance perception from reality. But often, we see what we expect to see.
Why does this matter to the Army? To the concept of “thinking like a soldier”?
Two weeks ago was the Maneuver Conference, where the focus of discussions was how the US would fare against a near-peer competitor, specifically Russia. During one of the presentations, a student in the Captains Career Course asked BG Daugherty, currently serving as Deputy Chief of Staff, G- 3, U.S. Army Europe, about the difficulty in distinguishing vehicles belonging to our allies versus the Russians, and the potential for fratricide. The answer was that the potential was high, because too many of our partners do not have Blue Force Tracking to be able to develop a common operating picture, but that with a lot of training, our soldiers would learn to recognize the difference, and that “visual discrimination” would be adequate for making decisions regarding whether to engage a potential target.
Ummm…. Well…. No….
In stressful situations (you know, like war, especially in the throes of a battle), what we see and what we perceive are not the same thing. So, even if soldiers could be trained to visually distinguish, for instance, Russian and Ukrainian military vehicles (some of which are produced in the same facility, and are actually the same vehicle), during times of crisis, they will all look the same. Soldiers will perceive a threat or an ally based on many things other than the “reality” of what they see, and this is not something that can be changed by training soldiers what the different vehicles look like.
After all, I can guarantee you that every person involved in the Charlotte incident can distinguish between a gun and book.
So what can be done?
The answer is two-fold, and training is part of it – but not as simplistic as “learning to visually distinguish”. The first part is a new, updated Identify Friend or Foe (IFF) system – one that does not rely on Blue Force Tracking, satellites, or a common operating picture, none of which works in a degraded environment. The second part is to fully incorporate that system’s use into the training environment.
In the 1990s, the Joint Combat Identification Panels – then called the Quick Fix Panels – were developed and deployed in under 18 months to prevent us from using our very lethal weapons on our own soldiers (or, more accurately, on their vehicles). My understanding is that the size and shape of the panels, which remain the only viable ground-based combat identification system (there are other systems available for planes and ships), cannot be used on our European partners’ vehicles due to their size and shape. Furthermore, the panels were called the Quick Fix because they were only supposed to be used for 2-3 years, with the developer assuming that our near-peer adversaries would be able to use them against us (as positive hostile identification) within that time period.
So, step one is a new IFF system. The system must be quick and easy-to-use, and a point-to-point system like the panels, so they continue to work even if satellite systems are unavailable. It must be available for all vehicles, not only the US platforms.
Step two is to train with this system. Soldiers who are making the fire/no-fire decisions must be experts at using the system, and have trained, to the point of muscle memory, how to do it and respond to the friend/foe distinction. Even then, there may be some perception issues, but this is the way to reduce it as much as possible.
Police officers do not get the benefit of using combat identification to differentiate bad guys; while fighting non-uniformed adversaries in Afghanistan and the Middle East since 2001, our soldiers haven’t either. But in a near-peer war, we not only can provide our soldiers with the capability, but we owe it to them to provide it and train to use it.
Especially in a world where someone is taping everything, and perception and reality are not the same, we owe it to our soldiers to provide them with an answer beyond “training”. As someone who believes in training, who believes that a soldier’s job, when not deployed, should be to train, it is vital that we understand the limitations of training, and base not only training, but also expectations and materiel development based on that understanding.
You can train muscle memory. You cannot train perception.
I am a theater geek. I love musicals. And road trips are made for listening to soundtrack after soundtrack and belting out tunes.
Last week, my dogs and I decided to check out Savannah (OK, I decided to check out Savannah, and they came along for the trip). With 9 hours in the car each way (thanks traffic!) I had a lot of time to think, listen, and sing.
For a while now, I’ve been thinking about how to expand John Sweller’s instructional design Cognitive Load Theory to apply it in a more global manner (beyond instructional design/education). Many hours of driving later, I think I’ve come across an easy multi-media way to explain it.
Under Sweller’s Cogntive Load Theory, if material is presented in such a way that the lesson’s load is too high or too low, the educational objectives will not be met: too low, and the information isn’t engaging enough; too high, and processing stops. Cognitive load of the material is based on intrinsic load (how complicated is the material), extrinsic load (how much mental power is needed because of how the material is presented), and germane load (how much effort is the individual putting into learning the material). [There’s also a whole part to the theory about Working Memory, and how, when demand on working memory exceeds capacity of working memory, processing slows or stops, resulting in cognitive overload. But I will save that for another post.]
What does any of this have to do with musicals?
The musical Hamilton, written by Lin-Manuel Miranda, is the hottest ticket, by far, in New York City right now. For one show in June, tickets were going for thousands of dollars each. When tickets go on sale, Ticketmaster’s website cannot keep up and they are all gone almost immediately. Why? It is a musical about a Founding Father. It is based on an 800+ page book by Ron Chernow, which never received nearly the amount of acclaim that Miranda’s musical version has.
From a cognitive load theory standpoint, the musical is far more likely to educate than the book is.
Leaving aside the artistic license taken by Miranda, viewers of the musical are much more likely to retain a greater portion of the material than readers of the book. Why is this? It is the same story (though the book is significantly more detailed); the intrinsic load of the material in both forms of presentation are approximately the same. We will pretend here that people who see the musical or read the book are equally interested in learning the material (though this is very likely a faulty assumption; there are many other reasons to go see a popular musical, and very few other reasons to invest the time into an 800 page book), making the germane load approximately equal.* The main difference, then, is the extrinsic load.
The musical is a modern translation of Hamilton’s life. Rather than concerning himself with replicating Hamilton’s tone or the culture of the day, Miranda uses hip hop, rap, and slang to tell the story. Conversely, Chernow uses direct quotes from the individuals involved and complex sentence structure and vocabulary throughout, generally targeting a much narrower audience. Anyone can watch Miranda’s telling and learn something; Chernow, with 800+ pages of text and footnotes, and 36 hours for the Audible version, requires far more focus to understand, learn, and retain the information.
The musical combines the first two chapters – 34 pages, or an hour and forty-five minutes in the Audible version – into a 4 minute song, performed this year at the Grammy’s:
It’s pretty easy to understand why 4 minutes would be easier to remember than almost two hours. But is it just the bevity? Is it just the addition of music?
No. It’s also the choice of language.
Unless you are reading Shakespeare (and I was going to use an example here from the Shakespeare spoof Something Rotten, also now on Broadway, but it seemed to get too convoluted), the exact words chosen to express a sentiment can either provide confusion or clarity. Using a $10 word when a 2¢ word will do unnecessarily increases the extrinsic load of the information. Using extra multi-media effects – such as fade-ins and outs in a PowerPoint presentation – that do not add to the story has the same effect.
At all times, it is essential to remember the goal of presenting information: if it is to inform, it is vital to properly balance intrinsic and extrinsic load (and, as much as possible, germane load). This is the case whether the goal is to educate or to provide soldiers with vital intelligence information. Poor information design contributing to cognitive overload is just as bad as, or worse than, too little information. And has long-term effects on each individual’s cognitive health.
* Someone with very low germane load – no interest in learning the material – will not learn it, even in the musical form, because they are, essentially, choosing not to. Nothing in the design or presentation of material can overcome the barriers an individual can put up to decide not to learn.
I have a bit of an Amazon.com addiction. I’m rather embarrassed to admit how much time I spend on the site, and how little impulse control I have when I’m there. Too often, I buy multiple books at once and then never read any of them. This year, I decided I would actually read (most of) the books I buy, and, to my credit, I have read a lot more (and bought a lot less) than in past years.
Since 2016 is half over, Amazon spent last week sending emails about best sellers in the first half of the year. Of course, I looked to see what made the various lists, and was surprised to see that I’ve read so many on the Business list. Not because I don’t read business books, and certainly not because I haven’t been reading, but because I didn’t think of any of these books as “business books”. I had read them in my pursuit to better understand the state of the research for what the Army calls the “Human Dimension”.
Seeing their classification, and that they are all doing so well made me start thinking about the common threads among the books. I decided to do a more structured (but still very unscientific) analysis of four of the books, with a fifth (released in December 2015) thrown in for good measure:
|  |
I summarized each chapter on to one 3”x3” Post-It Note, and then transferred the information from those, one thought at a time, to smaller ones that could be moved around, and put them all on a big piece of easel pad paper.
Once I moved them around, what did I find?
 |  |
Well, honestly, not as much as I thought. The things that struck me as being related when I thought about the books abstractly were clearly NOT key topics of given chapters, so didn’t make it into the analysis. Perhaps if I’d tried to link them serially, I would have seen more, but just grouping wasn’t as dramatic as I expected.
But that doesn’t mean there weren’t trends.
First: Originals had the least overlap with the other four books. That doesn’t make the information better or worse, just different. Peak and Grit had the most overlap (Duckworth actually references Ericsson’s work, so that wasn’t too surprising).
Now, for the important part: the trends.
The only semi-conflicting information was on the value of commitment based organizations – those organizations that are very focused on taking care of its employees as one of its prime missions. On the one hand, they tend to be very stable, and last longer in environments like Silicon Valley, despite the fact that such values were deemed “dead” in the 1990s. On the other hand, they tend to attract like-minded people, leading to groupthink and too little ability to adapt. Dissension in the ranks can be good (as long as there is psychological safety, according to a third book).
While the trends aren’t as strong as I expected, I would say that, looking at the totality of the information in the books: training is important. (OK, that comes as no surprise to anyone who knows me.) But there is more to it, especially for the Army. Taking each trend above:
There is a lot of research being conducted in many different arenas related to the “Human Dimension”. And with the number of people in the Army organization – 1M in uniform, 330k civilian – it is, without a doubt, easier to focus on materiel development than on the human dimension. But the human element – fixing training, improving recruiting metrics, understanding why there are so many impulse control issues among returning soldiers, and so on – is significantly less expensive (means) and provides the Army a way to significantly increase the readiness (ends) in the near-term.
First, I need to acknowledge something: this whole “remembering to blog weekly thing” is harder (for me) than it should be. So, no more scheduling. Of any kind. Though I will (probably) go back in and fill in some holes. Especially about Peak, a book that is somewhat relevant to today’s post (at least in my mind), but I still haven’t managed to put together my thoughts in a sufficiently coherent manner, so I’m not ready to push “publish” on last week’s post quite yet.
For a number of reasons, this year, I haven’t been able to travel the way I have in years past. So instead, I’m going to events at various DC think tanks. It’s that, or I never leave my house. As an introvert, I’m pretty comfortable with that, but it’s not healthy, so… think tank events it is. Last week’s was “Army Readiness, Fight tonight and Fit for Tomorrow” at the Center for New American Studies. The speakers were Mr. Daniel Feehan, Principal Deputy Assistant Secretary of Defense (Readiness); MG Walter E. Piatt, Director of Operations, Readiness and Mobilization, Army G3/5/7; Mr. Andrew Hill, U.S. Army War College; and Ms. Katherine Kidder, Center for a New American Security.
Readiness is GEN Milley’s top priority and, as he says “there is no other number 1”. Readiness is the ability to defeat one peer competitor while holding back a second without losing ground in the war on terror. That’s a lot to expect from an Army that has 14 years of continuous asymmetric warfare under its belt; no full spectrum operations in that time (and very little training for it); and soldiers who are tired from the continuous cycle of deployments. rip
Readiness is usually seen as a tri-pod; manning, equipping, and training must balance for the organization to be ready. For the Army, manning levels are set by someone else. (Though there is a lot the Army can do to effect its manning by having the right people – recruitment and retention – that’s not today’s topic.) Equipping? Soldiers have a lot of good equipment – thanks to 14 years of warfare – and what they don’t have now, they won’t have any time soon, based on development timelines. (In fact, in February at New America’s Future of War Conference – I told you, I’m going to a lot of think tank events – GEN Milley said he doesn’t see any major new systems coming, just upgrades to legacy materiel, for the next 5-10 years.) So the key to improving readiness, it seems to me, is training.
Training systems – whether live, virtual, or constructive – are expensive and cumbersome. They all introduce various levels of non-realism. And all are limited, either by the number of people who can use them at once, by the echelon that is going to get priority, or by the terrain they cover. Many are ‘better than nothing’ – but often, the juice (cost and what it takes to make them work) isn’t worth the squeeze (the learning achieved). And our soldiers (and taxpayers) deserve better than this. And readiness demands it. Time at CTC, the premiere training event, has been shown to be more effective for organizations that have had more pre-CTC homestation training. So how do we get more out there?
Complaints about the current training systems are plentiful.
Live training (e.g., MILES)
Virtual training (e.g., DSTS)
Constructive training
None of these concerns are new or unknown to the Army. PEO STRI is working to field new systems, but the process is so slow that it will be 2025 before there is a new individual MILES system fielded, and 2028 before an integrated live, virtual, constructive, and gaming system is fielded.
But what about the next 10 years? What about readiness now? How do we increase full-spectrum operations training?
I’ve spent almost a decade working this issue. It can actually be resolved fairly easily, but there’s actually no path through the system for an easy training system. That being said, here are my observations:
Most organizations – governmental (PEO STRI and the USMC both have MILES issues, and both are looking at them the same way) and other companies – look at the issues with MILES as primarily an issue with the interrogator. If the laser could penetrate better, the system would work better. To that end, they’ve considered other wavelengths (1550 nm vs the current 908 nm), other technologies (e.g., geopairing), etc, for the next generation systems – and wind up back with the old system (partly because of backwards compatibility issues that STRI can’t seem to break free from).
When we first started designing TIS, the Training and Identification System, we didn’t focus on components. We looked at SYSTEM requirements. Because of that, we have a very different solution. Rather than fixing the issues at the interrogator end, we were free to look at the whole issue. And that’s what we did. By making one addition to the MILES communication code and adding a COTS smartphone (in place of any other kind of radio) on the target, we can resolve three of the four complaints about MILES. (Details to follow in a subsequent post – or contact me for details.)
But we didn’t stop there.
Because we know that the long-term goal is to incorporate live training with virtual, constructive, and gaming technologies, we wanted to make sure that TIS could do that. So we incorporated plans for alternative means of firing the laser (current MILES technology “fires” off the force of the blank leaving the barrel – this could be very dangerous in an LVCG environment where the shooter may be facing something or someone other than what he is seeing), designs for incorporating the shot into the virtual world, and so on – again, using a COTS smartphone as one of the major components.
Training can be modernized – training MUST BE modernized in a much shorter timeline – to improve readiness. Homestation training can be available without expensive instrumentation (that requires environmental assessments and approvals), to multiple groups, without need for a lot of extra space or people to run the systems. Guard units can actually make full use of their 48 Unit Training Assemblies and at 15 days of Annual Training rather than spending most of the time trying to get the gear to work (if they can use it at their armories at all).
Realistic, affordable, useable training systems in the hands of soldiers is the key to increasing readiness. But it requires a means other than the traditional process, or it will be late into the next decade before soldiers have it.
This past week, I read the book Peak: Secrets from the New Science of Expertise by Anders Ericsson and Robert Pool. I had planned to use it as the basis for this week’s blog post (and I was actually going to post on time for once!) but there’s just too much to it, so it will have to wait. Suffice it to say, at least for now, if you’re looking for a great book on the Human Dimension, look at Peak.
So instead, tonight, I am going to reflect on metrics and the Human Dimension. This week, I went down to Columbus, GA for the Maneuver Center of Excellence Industry Day. It was a great opportunity to hear from the new Commanding General, BG Wesley, and the Commandant of the Infantry School, BG Jones.
It also offered an opportunity to talk to a number of people, both from the MCOE and other contractors, about the human dimension program. I was surprised to hear, on both sides, the issues people are having in attaching value to aspects of the program. In one case, it was in valuation of training. Actually, “surprised” is an understatement.
How do you value a new piece of hardware? You look at the increase in protection, the cost (or increase in cost) of the new equipment, and the costs associated with failure, and do fairly basic arithmetic. How do you value the human dimension program? It’s really not all that different. It’s that we aren’t used to thinking about it.
I have proposed several human dimension programs. All have associated costs. In some cases, as a contractor, I cannot access all of the numbers, but can provide some of them. Below, I will explain how I do this. I am not including my costs, as that information is company proprietary.
Program 1
In program 1, I ask the question whether any of certain personality and/or cognitive traits, if added to the ASVAB, would do a better job of selecting recruits in to the Army and helping them select MOSs for which they are suited. Today, the Army recruits approximately 60,000 soldiers each year; more than 1/3 do not complete their initial term. The numbers, as broken down by Brad Carson at CSIS earlier this week, is that 10-12% sign up but do not ever make it onto the bus for boot camp; 12-15% do not complete boot camp; and another 12-15% exit the Army at some point before the end of their initial term, for any number of reasons.
Basic training costs between $54k and $72k per soldier, depending on whether the MOS has One Station Unit Training or the soldier has to move between Basic and Advanced Individual Training. For ease, we’re going to assume an average of $63k per person. Of the 60k who start the Army in a year, and using the lowest numbers provided by Mr. Carson, 6k do not make it on the bus, 7200 do not complete boot camp, and another 7200 do not complete the remainder of their initial term. Ignoring the costs to recruit the 6k who don’t make it on the bus, just over 14.4k are still expenses to the American taxpayer, to the tune of almost $1B just for the cost of Basic Training. That does not include any additional costs the person may incur – including potential on-going medical costs through the VA.
The goal of the program is not to reduce initial term attrition to zero; it is to reduce the recruiting goal to 50k. This would save money for a lot of reasons, including time and effort expended by recruiters, and might reduce the number of recruiters needed (opening up spots in the reduced Army for other jobs). There will be cost savings in terms of training, and materiel, and salary, and benefits. And all of these are hard to quantify without more information than I can access. But I do know that reducing the recruits by 10k at 63k each will save $630M each year. Even reducing by 7200 (half of the individuals who don’t finish boot camp, and half that don’t make it through initial term) would be $454M PER YEAR.
Because my study is a very small fraction of this, and is not necessarily an on-going cost (though certainly could be as more metrics are developed and soldiering has new, different needs), it saves money.
Program 2
Program 2 asks the question of whether brain training programs transfer to real-world benefit to the trainee: do they do anything other than make the trainee better at playing the video game?
This is a significantly more complex program, but it’s also significantly shorter. Unfortunately, complexity = expensive, making it a hard sell.
Except that DoD (including the Department of the Army, but not only Army) continues to fund millions of dollars worth of studies on brain training – and none of them are looking at transfer. The successful completion of Program 2 would provide a template for testing transfer, and allow the research organizations to spend less money on basic research without any evidence of transfer. Unfortunately, I cannot get any current numbers of the money still being spent on brain training programs by DoD to provide that end of the equation.
Program 3
Program 3 is not mine; it’s a training program I heard about the other night. It was developed by a Large Contractor at the cost of $1M (in Independent Research & Development (IR&D) money), and was never completed or sold because of this metric issue: they – both the company and the Army customers – couldn’t figure out how to properly demonstrate its worth.
A single CTC rotation costs $30M for 1 month. Therefore, being well prepared for the rotation is critical. The value of this training would have been sending soldiers more prepared for the rotation. Better preparation could be demonstrated before going by showing improved scores on basic tasks. It could also be demonstrated after the fact by improved performance at the CTC.
Again, the numbers are not as clean as in Program 1, but that’s in large part because I am not familiar with how much the company was charging, how much increase in performance was expected, etc. But clearly, with CTC being so important, organizations invest in preparation. Comparing those costs are the key here.
In conclusion….
Many people still seem to think that the whole idea of the Human Dimension Program is grounded in “touchy feely stuff”. It is not. It is grounded in science – both hard and, yes, sometimes social science. There are real, true costs to implementing these changes, and real, true costs to the status quo. But both sides of the equation can have valuation attached. The costs may not be as easy to find (at least for contractors) as those for materiel, but they are available to those who can go look for them.
© 2016, Analysis First LLC