Towing readers with a section intro

Towing readers with a section intro

If you can hook readers right at the beginning of a section, that's amazing!  All that remains is towing them gently to the end.  

 

How do you do this without losing them on the way?  Well, here's a quick demonstration. Take this brief introduction to Section 2 in a paper:  

 

We conducted numerical experiments to study two questions:
(1) How do multiple stop itineraries affect revenue classes?
(2) Compared to current CPT measures, how much better does the policy perform?

 

 

Fine.  Now, what are you expecting next? 

 

If you're like other readers, you probably want answers to the questions the writer posed. Even if you didn't particularly care about those answers beforehand, as you read, you probably wondered, "OK, how do they?" or "How much better does it perform?"  

 

Well, let's pick up where we left off and find out...

 

In our first set of experiments, we randomly generated 24 problem instances of varying complexity, having time horizons τ in the set {20, 50, 100, 200, 500, 1000}. We considered a railway servicing L locations out of a single hub, where L∈{2, 5, 10, 20}.  This is an important, basic network structure of revenue management problems found in the railway industry. Each location l ∈ 1, ...., L is associated with two legs, i.e., to and from the hub, each departing at the end of the time horizon T. Hence, there are m = 2L resources. There are m single-leg itineraries and L · (L − 1) two-leg itineraries. 

We generated the revenue for each low-fare class from a discrete uniform distribution on the interval {15, 49}, and set the high fare for the same itinerary equal to five times that of the low fare.For simplicity, we considered stationary demand with the probability 0.2 for having no customer arrival in a period, and the other probabilities generated randomly. For each itinerary, we split the demand so that 25% was for the high-fare class, and 75% was for the low- fare class. 

In Table 3, we report the CPT seconds required to solve (P1) guaranteed to within Ω  = 5%, which is a practical tolerance given the data uncertainty encountered in the real world.

 

200 words later, you still don't have an answer.  Instead you've been bombarded by a host of technical details that, at this very moment, you weren't really interested in.  

 

And that's how the writer, who had managed to hook you, lost you.

 

Sometimes, towing a hooked reader is as simple as knowing what expectations we create and fulfilling them.  As long as we do that, our readers will likely come along for the ride.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Post by Varanya Chaubey
Image by kallerna (Own work) [CC BY-SA 3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

 

 

What Fawlty Towers teaches us about a Lit Review

What Fawlty Towers teaches us about a Lit Review

A paragraph map

A paragraph map