Trainsplit survey

[infobleep]

Then just leave the survey alone.

But my point is, it is then missing out a section of people and surely they want as many views as possible.

 

[superkopite]

But my point is, it is then missing out a section of people and surely they want as many views as possible.

I am sure its probably a very small cohort of people, possibly only yourself

 

[trainophile]

I must admit that there will be reasons other than price why you might reject one of the suggested alternatives. For instance there’s nothing wrong with Liverpool but I can get there in an hour for a day trip from my place in Southport, so I wouldn't ever select it for a stay.

 

[punxsutawny]

But my point is, it is then missing out a section of people and surely they want as many views as possible.

I am pretty sure most people understood the intended meaning of the question.

 

[ainsworth74]

I can't though as that question is mandatory. If I lie and I give you rubbish information, which might then lead to my other responses being considered someone playing around and discounted

Can you not just provide a rough estimate? That's what I did for the question as I'm not sure how many business trips I take in a year but I was able to have a guess (20 in my case) but there's no way that that is a completely accurate answer but I'm sure it's accurate enough for my answer to be relevant to whatever processing is being done to the survey responses

 

[duncombec]

That possibly wins the award for the worst designed survey I've taken this year. As well as being long and unwieldy, it seems to be set on the premise that people don't really have fixed intentions and will just decide to go somewhere else in the country on a whim - do people usually book train tickets before their hotel, for example?

- It was hard to know whether "coach" also included "bus", which may have been possible for some of the suggestions I got, and would be a viable alternative for a good portion of my leisure journeys.

- Who picked the list of destinations? Most were notable tourist places, but... Sheerness-on-Sea?! The Lake District... Edinburgh Castle... Newquay Beach... Sheerness Tesco? (Yes, I know it was probably picked as the nearest station to Leysdown, but stark contract to the more obviously touristic nature of the other locations).

- I gave my sample journey to Cardiff, Tuesday to Friday. The fact the algorithm would suggest I'd want to spend six nights in Edinburgh, or four nights in Newquay, or go to Cardiff from Sunday to Wednesday seems slightly bonkers.

- Similarly, in regard to the train ticket + accommodation questions, who would plausibly spend 6 hours in each direction to Edinburgh and back for a one night stay? What exactly would you expect to see between Sunday morning out and Monday morning back?

I can't though as that question is mandatory. If I lie and I give you rubbish information, which might then lead to my other responses being considered someone playing around and discounted


Possibly but it asks for the number of journeys as opposed to the approximate number. It's not my job to intwrpt the meaning of their question.

Even then I'm not sure how confident I would be with my estimate.

I know I take zero business trips a year, so for leisure, I entered 50. If each single direction counts, that's 25 trips a year, or approximately one every other week. In the summer, that frequency is probably about right, other times I might go somewhere in the evening, there will be a holiday included, so maybe a little high, maybe a little low... but anywhere between 40 and 60 would be reasonable (20-30 trips). I know I'll have taken more than 10 return trips so 20 would be too low, and I'd be fairly certain I haven't taken 100 return trips (so 200 would be too high).

 

[redreni]

It might have been an idea to tell potential respondents that the survey is aimed specifically at those with flexibility over dates and times.

Much of my leisure travel budget goes on football away days where I am, fairly obviously, constrained by the scheduling of the match. I may travel an hour earlier or later than the optimum train if it's a lot cheaper, but that's about it. I believe LNER regards an hour either way as 'flexibility' for the purposes of its offer to passengers, but I guess I would still be regarded by the rail industry as a fairly inflexible passenger and a lot of the questions weren't really relevant to me.

For a longer trip to a 3pm Saturday game (which is when they nearly always are in non-league), I will consider stopovers on the Friday or Saturday night, particularly where the saving in train fares compared with the day trip offsets at least to some extent the cost of the hotel, but a lot of the time the survey was asking me to guess if I would book a journey at a certain price without knowing the cost of the alternatives and often, without knowing the cost of stopping over. And for a great deal of the time it was asking repeatedly if I'll travel midweek, at different price points, which obviously is always going to be a 'no' in my case.

 

[Vexed]

I'm going to go against the grain here and say I didn't find it too bad. Even if you say none of the alternative times are suitable for nearly/all answers it's still good data. If

If only people who can take alternatives answer the survey then it's not representative and won't help with what I think the point in this is - to prove that with a (as close as you can easily get) representative sample a decent chunk are flexible and would benefit from seeing advance quota availability over a longer period. It did get quite repetitive after a while though.

 

[infobleep]

Can you not just provide a rough estimate? That's what I did for the question as I'm not sure how many business trips I take in a year but I was able to have a guess (20 in my case) but there's no way that that is a completely accurate answer but I'm sure it's accurate enough for my answer to be relevant to whatever processing is being done to the survey responses

I suppose I could. I just like to be accurate.

I am pretty sure most people understood the intended meaning of the question.

But then some people understand you can't use an operator only ticket on another operators train even if staff direct you to take the next train, which I'd run by a different operator.

But not everyone gets this.

 

[sheff1]

Similarly, in regard to the train ticket + accommodation questions, who would plausibly spend 6 hours in each direction to Edinburgh and back for a one night stay?

I have travelled 6 hours to Aberdeen and back for a one night stay & 4/5 hours to various other places.

Wouldn’t expect to be offered Aberdeen (or Edinburgh) though if I had searched for trip to Cardiff!

 

[Haywain]

Similarly, in regard to the train ticket + accommodation questions, who would plausibly spend 6 hours in each direction to Edinburgh and back for a one night stay? What exactly would you expect to see between Sunday morning out and Monday morning back?

One night stay? Ive done journeys like that for day trips loads of times.

 

[35B]

Similarly, in regard to the train ticket + accommodation questions, who would plausibly spend 6 hours in each direction to Edinburgh and back for a one night stay? What exactly would you expect to see between Sunday morning out and Monday morning back?

My wife and daughter did a month ago, to go to a concert. Travel up, check in, eat, go to concert, sleep, check out, travel home.

Just because I would not do x does not mean others won’t do x, and surveys are designed to show how many may do x.

Perhaps the more interesting question is the underlying assumption in the survey about how people plan travel. It had never occurred to me to choose which days I’d travel based on ticket availability; the trip is usually set around other constraints and then I’m booking travel within those constraints.

 

[redreni]

It had never occurred to me to choose which days I’d travel based on ticket availability; the trip is usually set around other constraints and then I’m booking travel within those constraints.

Thanks for saying so much more succinctly what I said further up.

We're not supposed to have other constraints, of course. The whole national transport strategy seems to be predicated on avoiding the need to build significant new infrastructure by using price to smooth out demand, thus getting more journeys out of existing capacity.

Our other constraints don't really fit into that theory so I strongly suspect it's not career-enhancing within DfT or Treasury to talk about them too much. In my case the other constraints mean they might succeed in pricing me off the busiest trains, but only onto other modes, notably onto major roads which also lack the capacity to cater to peak demand.

 

[duncombec]

I have travelled 6 hours to Aberdeen and back for a one night stay & 4/5 hours to various other places.

One night stay? Ive done journeys like that for day trips loads of times.

My wife and daughter did a month ago, to go to a concert. Travel up, check in, eat, go to concert, sleep, check out, travel home.

Just because I would not do x does not mean others won’t do x, and surveys are designed to show how many may do x.

Yes... I thought about concerts and the like after I sent the message. I was definitely comparing it against my original search for a "leisure trip" of three nights in Cardiff, and wondering how much leisure I was having (in the sense of city break) in a few hours on Sunday evening!

Perhaps the more interesting question is the underlying assumption in the survey about how people plan travel. It had never occurred to me to choose which days I’d travel based on ticket availability; the trip is usually set around other constraints and then I’m booking travel within those constraints.

I think that's true for a lot of people, and perhaps why I found the premise of the survey a little unusual. That said, I did recently change a flight time from my intended "about lunchtime" to "evening" because, on booking, the evening flight time was so significantly cheaper that it was worth me also avoiding peak hour public transport fares into London in the morning to reach the airport. But it was still the same day, and 4-5 hours difference - not a wildly different number of days or to a completely different destination.

I can see the possible benefits of a system such as this, though - a "micro push", as it were, to consider close alternatives to your inputs. E.g. "Your desired departure time of 8am is a Premium Super-Duper Bells and Whistles Ticket, but you can save 25% by travelling at 9am on a Slightly Higher Than Average Flexible Ticket, or 35% at 10am on a Cattle Class Express that actually gets you there 15 minutes faster than the 9am so you're not even losing the whole hour" sort of thing, that might be too far down National Rail's time span but for the "time rich" may not actually create too much of a dent in their schedule.

 

[nwales58]

Perhaps the more interesting question is the underlying assumption in the survey about how people plan travel.

Although the survey ignores that (it rules out most people's preferred option of a flexible walk-up ticket, reasonably priced for instance), that *is* the correct design for the problem in hand:

Our goal is to estimate potential revenue growth if we and other retailers gain access to an availability feed from RARS - ADS ... this access would allow us to display a week's worth of availability ... making it easier for customers with flexible travel plans to find quieter and cheaper options.

In other words: out of people who want to trade time/date and/or destination against cost, how elastic are those choices against cost? Data from those who do not trade does not affect the parameters of the model being fitted for this problem.

This is different from a more usual sample survey (e.g. to find what %ge of people would trade time/date against cost).

It is also why the self-selected sample is not a flaw (unusually - most internet surveys are statistically near-useless).