Learning curve

[Larry Browning]

Random thoughts.

The term "Steep learning curve" meaning taking a long time to learn something seems completely backwards to me. In my mind a steep learning curve would mean it was quick and easy to learn while a shallow learning curve would mean slow and difficult.
If you imagine a line graph with the vertical axis representing what is learned and the horizontal axis representing time. The line would be very steep when you learned something quickly, while it would be shallow when you learned very little over a long period of time.

What do you think?

[Dan Hintz]

If the point at which you know enough to be productive is the top of a hill (of knowledge), a "steep learning curve" means you have a very large hill to climb, i.e., it takes a LOT of knowledge to get there.

Walking 100' is a lot easier than climbing 100', so the X-axis shouldn't be viewed as time, per se...

[David Ragan]

If the point at which you know enough to be productive is the top of a hill (of knowledge), a "steep learning curve" means you have a very large hill to climb, i.e., it takes a LOT of knowledge to get there.

Walking 100' is a lot easier than climbing 100', so the X-axis shouldn't be viewed as time, per se...

Well, the time factor is the world we live in. How about effort being on the X-axis?


Random thoughts.

The term "Steep learning curve" meaning taking a long time to learn something seems completely backwards to me. In my mind a steep learning curve would mean it was quick and easy to learn while a shallow learning curve would mean slow and difficult.
If you imagine a line graph with the vertical axis representing what is learned and the horizontal axis representing time. The line would be very steep when you learned something quickly, while it would be shallow when you learned very little over a long period of time.

What do you think?

I am so glad you brought this up Larry.

This concept has confounded me for some time. I agree completely. It is probably put forth by the same folks who use apostrophes, when a plural is meant; and who say "an historical moment"; and folks who water their lawns in the evening (as opposed to the early AM), and those who bang up your tools and think nothing of it, and-those parents who think it is the school's responsibility to make their kids into good citizens.

[Larry Browning]

If the point at which you know enough to be productive is the top of a hill (of knowledge), a "steep learning curve" means you have a very large hill to climb, i.e., it takes a LOT of knowledge to get there.


To me the word "curve" implies a line graph and not an imaginary hill. No?

[Erik Loza]

I always took "steep learning curve" to mean that a person was given a whole lot of challenging information and expected to retain that in a very short amount of time. Such as starting a new job and having to master all sorts of new skills on the fly. A-la, "throwing someone in the deep end of the pool".

Erik

[Larry Browning]

I always took "steep learning curve" to mean that a person was given a whole lot of challenging information and expected to retain that in a very short amount of time. Such as starting a new job and having to master all sorts of new skills on the fly. A-la, "throwing someone in the deep end of the pool".

Erik
I think I am hung up on the word "curve". Maybe if the phrase were "steep hill to climb" it would make more sense. But if the word curve, implies a line graph (which it does to me), then it is a completely illogical phrase. Well, not illogical, I suppose, but rather backwards.

[ryan paulsen]

A line is just a curve where all of the points line up with a constant slope.

Think of x axis as time, y axis as the amount of skill/information needed to be obtained.

A process which required you to learn a large amount of skill over a short period of time would require a steep curve.

[Ken Fitzgerald]

I always thought a "steep learning curve" meant a tremendous amount of information to learn before one became knowledgeable about a specific topic or proficient at a task with no specific time factor involved.

[Larry Browning]

A line is just a curve where all of the points line up with a constant slope.

Think of x axis as time, y axis as the amount of skill/information needed to be obtained.

A process which required you to learn a large amount of skill over a short period of time would require a steep curve.
Ok, not being a mathematician, which is the horizontal axis and which is the vertical?

I have always visualized time as the horizontal and amount learned vertical. Are you saying reverse that?

lots learned
.....................|
.....................|
.....................|
.....................|
.....................|
knows nothing|______________________
......................time-->

So the line would be very steep the quicker and easier something was to learn and very shallow the harder it was to learn.


A process which required you to learn a large amount of skill over a short period of time would require a steep curve.
That's right, learning a lot over a short period of time would be the case if something was easy to learn, you would pick it up right away and therefore a steep learning curve. OTOH, if it took you a really long time to learn because it was a difficult concept, you wouldn't learn much over a long period of time, so that would be a shallow learning curve. I'm so glad you agree with me!

[Pat Barry]

If the point at which you know enough to be productive is the top of a hill (of knowledge), a "steep learning curve" means you have a very large hill to climb, i.e., it takes a LOT of knowledge to get there.

Walking 100' is a lot easier than climbing 100', so the X-axis shouldn't be viewed as time, per se...

This is the way I've always thought of it. Climbing a steep hill requires much more effort (steep learning curve) than a hill without so much slope (shallow learning curve).

[Larry Browning]

I always thought a "steep learning curve" meant a tremendous amount of information to learn before one became knowledgeable about a specific topic or proficient at a task with no specific time factor involved.


I have always associated time with the phrase. In fact you implied time by your use of the word "before". I do not know how the concept of time could not be involved. By saying something had to happen "before" something else, then time is involved and a key part of the concept.

[Larry Browning]

This is the way I've always thought of it. Climbing a steep hill requires much more effort (steep learning curve) than a hill without so much slope (shallow learning curve).


Where does the curve part come in?

[Ken Fitzgerald]

I always thought a "steep learning curve" meant a tremendous amount of information to learn before one became knowledgeable about a specific topic or proficient at a task with no specific time factor involved.


I have always associated time with the phrase. In fact you implied time by your use of the word "before". I do not know how the concept of time could not be involved. By saying something had to happen "before" something else, then time is involved and a key part of the concept.

I didn't say time wasn't involved. I said no specific time factor involved.

For example, due to different natural abilities, one person may learn something quicker than another to attain equal amounts of knowledge or skills. If something requires a certain amount of application and experience, one person may get to a higher level of proficiency quicker just because they have more time to apply to gaining the needed experience. The amount of applied time or the quality of the experience may cause the time to vary between individuals.

In the end, I thing the term is meant to be just a relative statement.

[John Lanciani]

http://3.bp.blogspot.com/-dgVJivuOk6o/T5dMu4D5YWI/AAAAAAAAAoI/bDsKWiV7QxE/s200/Steep-Learning-Curve.png





 (http://3.bp.blogspot.com/-dgVJivuOk6o/T5dMu4D5YWI/AAAAAAAAAoI/bDsKWiV7QxE/s1600/Steep-Learning-Curve.png)I always think of the phrase like this. Lots to learn in the beginning, becoming less over time.

[Larry Browning]

I didn't say time wasn't involved. I said no specific time factor involved.

For example, due to different natural abilities, one person may learn something quicker than another to attain equal amounts of knowledge or skills.

In the end, I thing the term is meant to be just a relative statement.

How does the word quicker not imply time?
What is the difference between time factor and time?

[Larry Browning]

http://3.bp.blogspot.com/-dgVJivuOk6o/T5dMu4D5YWI/AAAAAAAAAoI/bDsKWiV7QxE/s200/Steep-Learning-Curve.png





 (http://3.bp.blogspot.com/-dgVJivuOk6o/T5dMu4D5YWI/AAAAAAAAAoI/bDsKWiV7QxE/s1600/Steep-Learning-Curve.png)I always think of the phrase like this. Lots to learn in the beginning, becoming less over time.

Exactly, the steeper the curve, the faster you are learning.
By the way, the learning curve for me figuring out how to use a drawing program and posting the results on SMC is very shallow, in fact it might actually be flat:o

[Scott Shepherd]

There must be a steep learning curve on the learning curve ;)

[Larry Browning]

For example, due to different natural abilities, one person may learn something quicker than another to attain equal amounts of knowledge or skills. If something requires a certain amount of application and experience, one person may get to a higher level of proficiency quicker just because they have more time to apply to gaining the needed experience. The amount of applied time or the quality of the experience may cause the time to vary between individuals.

So what you are saying is that the learning curve is different for different people. I totally agree with this. I am just saying that, for me any way, when I say or think of a steep learning curve, I am thinking that it must be an easy concept, because it is quick to learn. This is what makes sense to me anyway.

[Larry Browning]

There must be a steep learning curve on the learning curve ;)


No, I think it is very very shallow:eek:
Lots of time going by and nobody is learning anything. But I am having fun talking about it.

[Lee Schierer]

Where does the curve part come in?

Just to cause even more confusion I found this on the internet:

"The first person to describe the learning curve was Hermann Ebbinghaus in 1885. His tests involved memorizing series of nonsense syllables, and recording the success over a number of trials. The translation does not use the term learning curve—but he presents diagrams of learning against trial number. He also notes that the score can decrease, or even oscillate.[5][6][7]

The first known use of the term learning curve is from 1909 : "Bryan and Harter (6) found in their study of the acquisition of the telegraphic language a learning curve which had the rapid rise at the beginning followed by a period of retardation, and was thus convex to the vertical axis."[2][5]

Psychologist Arthur Bills gave a more detailed description of learning curves in 1934. He also discussed the properties of different types of learning curves, such as negative acceleration, positive acceleration, plateaus, and ogive curves."

It sounds like your understanding of a steep learning curve is a factor in understanding what a steep learning curve is.

[Bruce Volden]

I see all the mentioning of the term "shallow".

Well, that works for me.

Bruce

[Phil Mueller]

I always understood it as a factor of learning and trial (or experiences). So for example, making a dovetail joint. If you are asked to make a perfect dovetail joint with just three attempts it's a steep learning requirement. OTOH, if you are required to make a perfect dovetail joint over 100 attempts, the learning curve is much less steep. Time isn't necessarily a factor, but often is.
The curve is related to the averages of many individuals, with # of dovetail attempts on one axis and quality of dovetail on the other...it's the average of many that makes the curve.
An individual's actual results may be a line that goes up and maybe even jogs down as quality varies from dovetail to dovetail until perfection is achieved.
It has come to express a lot to be perfected in a short period or number of "experiences"

[Larry Browning]

I think most people focus on the word "Steep" which, by itself implies more difficult than shallow. I seem to focus on the word "curve" which implies a line graph.

[charlie knighton]

the learning curve

the fallacy is there is no correct answer, just different styles

[Dan Hintz]

I really must learn to be more selective as to the discussions I join... ;)

[Brian W Smith]

Try using trajectory instead of curve.........and maybe incline vs steep.

[Pat Barry]

I think most people focus on the word "Steep" which, by itself implies more difficult than shallow. I seem to focus on the word "curve" which implies a line graph.

Sure Larry, a line graph, just as shown above. What is difficult to understand about that? That curve is the generally accepted 'learning curve' you see in all the text books. A steep learning curve means a lot to learn in a little time. The opposite is less to learn and maybe more time therefore the 'shallower' curve. But all things are not equal so with the steep learning curve there is a lot to learn therefore the asymptote of that steep learning curve is much higher than the shallow curve.

[Robert Engel]

OK guys this has gone on long enough.
I'll clear this whole thing up.

Whether its a curve or straight line, the analogy is climbing a hill, so "steep" refers to difficulty.

Different people get there at different times and some don't get there at all.
The angle or curve will be different for each one of us.

How long it takes to get to the top (learn it) depends on things like:

1. How fast you learn (i.e. how smart you are, how fast you learn from mistakes, etc.)
2. Natural talent.
3. How quickly you develop your skills.
4. How motivated you are.

We all know this and we should have better things to do, shouldn't we?

[Myk Rian]

when I say or think of a steep learning curve, I am thinking that it must be an easy concept, because it is quick to learn. This is what makes sense to me anyway.
That's backwards. It is meant to mean something is harder to learn.

[Charlie Velasquez]

Let me give it a shot. For illustrative purposes we will assume two different tasks to learn. Each task has the same number of different components (we'll say 5 for our example) and each component takes one minute to learn.
Task1: The components are disjoint. One component does not do much to affect the others.
Task2 is highly interactive between the components.

Lets learn task1 and "graph" the progress.
To learn component 1= 1 minute.
To get to stage 2 of the task the learner had to learn c1, then learn c2,=2 min.
To get to stage 3 of the task the learner had to learn c1, then c2, then c3=3 min and so on; so the graph would look like
Stage1: 1
Stage2: 1,2
stage3: 1,2,3
stage4: 1,2,3,4
stage5: 1,2,3,4,5
Pretty linear.

Now Task2,- to learn the first task it takes 1 minute.=1min, To learn the 2nd component it takes another minute, but there is an interaction between components 1 and 2 and you must learn that also; so you have a minute for c1, a minute for c2, and a minute for Interaction(12)= 3 minutes.
To learn component 3 you have the time you invested in stage2, c3, plus how c3 interacts with c1 ,c2,and I(12) =7min. And so on.
So the actual learning might look like:
Stage1: 1
Stage2: 1,2,12
Stage3: 1,2,3,12,13,23,123
Stage4: 1,2,3,4,12,13,14,23,24,34,123,124,134,234,1234
Stage5: 1,2,3,4,5,12,13,14,15,23,24,25,34,35,45,123,124,12 5,134,135,145,234,235,345,1234,1235,1345,2345,1234 5

And the graph might look like:
Stage1: 1
Stage2: 1,1,1
Stage3: 1,1,1,1,1,1,1
Stage4: 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1
Stage5: 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1

And you begin to see the curve. In other tasks there would be varying amounts of interactivity but you get the idea...

ps... larry, y axis is time..... Rotate my pictograph 90* counterclockwise.

[Charlie Velasquez]

Or... It could be that as you become more versed in the task it becomes easier to learn new components so the graph might look like this on our 5 component task
(0000=1 minute)
Stage1: 00000000 (2 minutes)
Stage2: 00000000,000000 (1.5 minutes for c2)
Stage3: 00000000,000000,0000
Stage4: 00000000,000000,0000,00
Stage5: 00000000,000000,0000,00,0


And it starts to flatten out some.


This could be compared to a task that might be easy to learn in the beginning, then get more complex as you go along. Target shooting comes to mind. To be good at ten feet is not as complicated as being good at 1000 yards.

[Larry Browning]

Sure Larry, a line graph, just as shown above. What is difficult to understand about that? That curve is the generally accepted 'learning curve' you see in all the text books. A steep learning curve means a lot to learn in a little time. The opposite is less to learn and maybe more time therefore the 'shallower' curve. But all things are not equal so with the steep learning curve there is a lot to learn therefore the asymptote of that steep learning curve is much higher than the shallow curve.
This absolutely supports my understanding of what "steep learning curve" means. If something is easy to learn your rate of learning is very fast and therefore generates a steep learning curve. If something is difficult and hard to grasp then your learning rate is slow and therefore generates a shallow learning curve.
It is not difficult to understand. I never said I didn't understand it, I said I don't understand why people use the term steep learning curve to describe something that is difficult to learn.

[Scott Shepherd]

This absolutely supports my understanding of what "steep learning curve" means. If something is easy to learn your rate of learning is very fast and therefore generates a steep learning curve. If something is difficult and hard to grasp then your learning rate is slow and therefore generates a shallow learning curve.
It is not difficult to understand. I never said I didn't understand it, I said I don't understand why people use the term steep learning curve to describe some thing that is difficult to learn.


Larry, I think that's right up there with people saying "I could care less" when they are trying to say how little they care about things. It should be "I couldn't care less", but most people say "I could care less". Once people start saying it's all over. Maybe that's what's happened here over time, it's been said incorrectly and now it's the norm.

[Larry Browning]

I really must learn to be more selective as to the discussions I join... ;)

Yeah! Next I will start a thread to discuss when clothes turn into laundry and then when laundry turns back into clothes again.

[Larry Browning]

Larry, I think that's right up there with people saying "I could care less" when they are trying to say how little they care about things. It should be "I couldn't care less", but most people say "I could care less". Once people start saying it's all over. Maybe that's what's happened here over time, it's been said incorrectly and now it's the norm.Exactly! This is a perfect analogy.

[Larry Browning]

I see all the mentioning of the term "shallow".

Well, that works for me.

Bruce
And your point is?.....

Well, of course this is a meaningless thread that is a total waste of time, but I find it an interesting discussion. This has never stopped me before, why should it now!

[Larry Browning]

OK guys this has gone on long enough.
I'll clear this whole thing up.

Whether its a curve or straight line, the analogy is climbing a hill, so "steep" refers to difficulty.


This simply supports my theory as to why people use the term incorrectly.

[Pat Barry]

This absolutely supports my understanding of what "steep learning curve" means. If something is easy to learn your rate of learning is very fast and therefore generates a steep learning curve. If something is difficult and hard to grasp then your learning rate is slow and therefore generates a shallow learning curve.
It is not difficult to understand. I never said I didn't understand it, I said I don't understand why people use the term steep learning curve to describe something that is difficult to learn.
No - you are not getting it

[Charlie Velasquez]

This simply supports my theory as to why people use the term incorrectly.I think it is used correctly. I think you have transposed the axes. I would consider the progression through learning a task as the independent variable (maybe demarcated as a percentage) and the time it takes as the dependent variable. Then the graphs will show a steep curve for the more complex and a less steep curve for the simpler, faster to learn tasks.

[Larry Browning]

No - you are not getting it


What am I missing then?

[Larry Browning]

Someone PLEASE clear something up for me. Is the X axis the vertical or horizontal axis?

[Myk Rian]

Example:
I am just starting to mess around with Arduinos. They use the C++ language, in most cases.
I know none of it. I also have lost much knowledge of electronics. Particularly IC circuits.
I have a lot to learn, and it is a steep slope/curve ahead of me.

[Larry Browning]

I think it is used correctly. I think you have transposed the axes. I would consider the progression through learning a task as the independent variable (maybe demarcated as a percentage) and the time it takes as the dependent variable. Then the graphs will show a steep curve for the more complex and a less steep curve for the simpler, faster to learn tasks.



Are you saying that time is the vertical axis? This does not make sense to me. Every line graph that I have ever seen that involves time has time as the horizontal axis.

[Larry Browning]

Someone PLEASE clear something up for me. Is the X axis the vertical or horizontal axis?
OK since no one will tell me, I googled it. X is horizontal and Y is vertical. (X to the left and Y to the sky)

[Larry Browning]

Example:
I am just starting to mess around with Arduinos. They use the C++ language, in most cases.
I know none of it. I also have lost much knowledge of electronics. Particularly IC circuits.
I have a lot to learn, and it is a steep slope/curve ahead of me.

So, your theory is that the term "steep learning curve" does not refer to a line graph at all, but to a hill or maybe a steep curve in the road? Hmm, interesting!

Or maybe the Y axis does not represent difficulty or rate of learning, but only amount to learn, so the Y axis gets longer(taller) based on how much there is to learn, not how difficult or fast it is to learn. Therefore, given a fixed amount of time, the slope gets steeper as the amount of learning needed increases. Hmmm.

[Mel Fulks]

Some stuff about it on line saying it has been used both ways and that current use is opposite of original meaning. I guess it is now interpreted by context, knowing the speaker,tone of voice, and hand signals.

[Charlie Velasquez]

Generally we put the independent variable on the x-axis and the dependent variable on the y. So, sometimes time goes on the y. The way you have it would answer the question, "How much can you learn in x amount of time?"

Instead, we want to answer the question, " How much time does it take you to learn x?"

[Pat Barry]

What am I missing then?


I'll chalk it up to the idea that you are a digital guy and this is an analog problem :)

[Scott Shepherd]

I thought it would be more along the lines of this shape curve Larry, meaning that on your first day or learning something new, there was only so much you could learn, because your knowledge of the subject limited you to understand so much that was ahead of you in the learning process. With each day, you get more educated on the subject matter, and the more educated you become, the more you realize that you don't know. For example, if you learn item #1 on day one, then item #2, #3 on the next day, maybe items #2,3,4,5.... all have 10 items of their own. Where you learned 1 item on the first day to get your knowledge, now, with the more items you are exposed to, the more items that you know less about, meaning you have to learn 10 times more on the 5th item as you did on item #1. It's like a family tree. You can start at one person, but then you learn about the immediate family, then the cousins, aunts, uncles, and then their families. Pretty quickly, there's 1000's of people in that tree. Same thing here, only with knowledge, meaning the more you are exposed to, the more you have to learn, thus a steep learning curve.

That's my take on it, but that could be totally wrong.

317051

[Larry Browning]

I'll chalk it up to the idea that you are a digital guy and this is an analog problem :)


That has GOT to be it!:eek:

[Larry Browning]

Generally we put the independent variable on the x-axis and the dependent variable on the y. So, sometimes time goes on the y. The way you have it would answer the question, "How much can you learn in x amount of time?"

Instead, we want to answer the question, " How much time does it take you to learn x?"
This makes sense sorta! I just have a hard time visualizing time being on the vertical axis.

[Robert Engel]

I think this whole thread is a steep curve leading to nowhere.......

[Larry Browning]

I think this whole thread is a steep curve leading to nowhere.......
Absolutely! Isn't it GREAT!
I have really enjoyed this thread. Sadly, I suppose there is nothing much else to discuss on the subject.
What I want to know now is, have I convinced anyone to start using the phrase "shallow learning curve" instead of "steep learning curve"?.......
I know that I haven't been convinced to change my ways!

[David Ragan]

I stick my original idea and failed hijack of this thread