e-Learning Media: Self made constraints : Part 1: Estimation Models

I was discussing a project with a long time colleague of mine from the same industry. We quickly started to rue the fact that e-Learning has been stereotyped to be a screen based activity.

Screens would contain either

1. Animation
2. Interactivity
3. Video
4. Simple text and Graphic.

These over a period of time, has got slotted into level 1,2,3,... for quick estimation and budgets. This, I always maintain is the wrong estimation technique. It is

1. Anti-pedagogy: It is the confluence of a chosen combination of these AS PER subject matter that defines the learning. Pre-configuring levels and fitting content into these levels is a dump of content in the determined areas.
2. Anti Pattern for User Experience: When you define levels, you say, in effect, that the course I am giving you is going to contain simple text and graphic with page turner, or screens which have a said interactivity and so on. This is not user experience. Where did we care for users and ask them what would they like to see and how they want to experience e-Learning ?
   
3. Opposed to the development teams real efforts: Levels define the output based estimates. But a delivery team takes in so much constraints, loose ends, additional but allied tasks, that they are not truly reflected in the estimates and hence the value of the entire project.

Still in my own team, I tend to join the chorus on discussing numbers with this model. So deeply entrenched is this culture, that the alternates are nowhere near main stream adoption.

Some alternates I have seen are:
1. Points based estimation: Have a unit for estimation and for various tasks, assign weighted points. These then sum up together multiplied by least effort value gives the total effort for which quote is given. This mimics functional point estimation in IT projects. However the unit figure typically is an abstract estimate, which needs more science for an effective measurement. It is best bet for approximation of efforts, does it give a framework for tracking efforts ?

2. Task based estimation: This technique is to list down the tasks in chronological order and rate the efforts against them. This is a safe bet and good estimation technique that allows for tracking as well. But the effectiveness of this depends on the granularity of defined tasks. Most of the time, we do so many out of the box tasks relevant to move the project that it seeps through the cracks of this estimation resulting in more variance.

3. Resource Loading based estimation: This at times really works. Though very crude and does not have scientific value, this gut estimation of resource loading over the project duration is normally close in measuring actuals. Yet, this measures efforts and cost of creating e-Learning. Does it measure the output and value of the package for e-Learning ?

4. Course-hour seat time based estimation: Suffers from same pitfalls as level based estimation, this is again anti thesis for pedagogy and andragogy. How can some one calculate the time I will remain in my seat for learning or studying ? May be reading. Yes. But is it e-Learning? What we talk here is only a notional time and I usually argue that it should be referred to with some good notation format and never as a value.

For example: SeatTime (Click Throughs):10 hours, SeatTime(Audio Length): 10 hours, SeatTime(Voice Reader): 10 hours.

These atleast suggest the various ways we arrive at the course hour and NEVER BY JUDGING LEARNING TIME.

Any other models, you use to effectively measure and track time of tasks, user feedback, value that is aligned with e-Learning and betters the above list. Please leave your comments.

Estimation - A near correct approach

Isnt it any company dream to have a correct estimate, correct price with normal profit and earn it upon delivery ?

Well, after much experience with simulation training models, we have intrepreted our experience to address this common problem. Infact the delivery team this time felt that overruns, over time and over work is killing their creativity. Hence came up with near to accurate tracking times and days required for their work.

Having emperical data is always a master stroke. Here it is a doubly sharp as data is accurate enough to be tracked and reported immediately with as many minor tasks that can be reasonably captured. The estimation technique has been refined with various baselines across multiple projects, that now our estimation technique of any simulation exercise is
= number of clicks determined*$ value for efforts.

Number of clicks is a very obvious number to measure during proposal stage. The value size is so granular that the % of assumption error will hardly be 5-10%. While traditional estimation models with the same error margin result in 15-20% normal variances, in this case, you can cover the variances in the cost, which is fraction less than normal costing models.

Our experience which is being field tested is hoped to confirm our postive thinking of acheiving 20% lesser cost and 15% less schedule time for delivery.

You wonder about the compromises made in the solution - Very minor which has not been enabled for want of time to deliver projects and not because of any technical limitation. However as my boss said today - you better revise your ID strategy for better communicating instructions than making them look machine instructed sounds.

Thanks to timely boss reminder that we are taking course correction to the ground principles that any e-Learning is not about technology but enable better visibility to Instruction principles through proper application of technical solutions.

Estimation - Unitization by solution or by components - Study of 2 Practices

Having been with e-Learning domain for around 9 years now, hence with little bit of confidence I can say that estimation techniques have been learnt the hard way and is still being perfected with various models. The maturity levels of gathering emperical data and improving upon the past estimation vs execution effort variances are rarely heard by me in my e-Learning friends circle.

Starting with per hour of learning model to Instruction design based effort multipliers to notional complexity level multipliers and currently looking at effort based estimation, I can say that e-Learning companies are still trying to get the "margin success" a recurring phenomenon.

In the company I work for is no different. Being an IT company delivering e-Learning content design services, it is expected that we too provide hueristic effort estimates like LOC, FP, resource loading chart (in case of a top-down approach) etc as with other technology projects. Any amount of convincing that we are different does not help either of us, since the sales team tuned to sell IT finds it complex to sell e-Learning. However few people who have had success with e-Learning find it an indispensable selling proposition for them to make some quick targets.

So initially we sailed in the same boat with classifying levels of interactive complexity and deriving best estimates by considering the good circumstances. Quick comparison with actual efforts we executed earlier confirmed our estimates are in right track.

Recently, we need to bid for an entire account where there was competition. The focus was submission of costs for various solution features. This broke our foundation with which we we defined levels - "based on interaction complexity efforts". This levels model required a great deal of documentation to make an inclusive list. The manager in me considers it to be a not good usage of effort.

While we mind mapped in our brain storming session, we found that there are various solution features, like PPT conversion, simple page turners, games, 2D Vector illustrations, 2D, 3D animations, 25 different interactivities and each of them can be combined to offer a solution in predictable ways.

The results were astonishing. Across the solutions there was a great overlap of efforts and it is indeed possible to group various solution in multiple price levels.

So instead of saying

1. "For a given price I will give you Flip book, Drag and Drop, Match the Following and MCSS interactivities", or
   
2. "My Instruction Designer would provide you with mandatory course features which is a passive animation and no support/interactive levels", I can now say

"For the given business requirement, I can think of the possible 2 solutions that address the need in the same price band. It is now dependent on the target users preference to choose one solution over another. So we would like to give you our User Research based Design Services."

Would you give your project to me, if I would tell you the above story ?
Well, the result of our submission to our client is still waiting for their consideration.