Great Western Coffee Shop

https://www.gazetteandherald.co.uk/news/17742410.customer-satisfaction-is-at-a-record-high-for-gwr/?ref=mr&lp=7

Not sure if this shouldn't be under "smoke & mirrors"  :D

Like all surveys, it depends what you ask and when you ask it. Is the questionnaire published, it should be for a reputable survey. What was the sample size, it really needs to be more than 1000 to have much statistical validity.

But who cares about all that it can still make a good (or bad) story, depending on the results.

It's not hard to find out from the National Rail Passenger Survey pages of Passenger Focus. That says:

There's a bit about methodology in the main report (https://www.transportfocus.org.uk/research-publications/publications/national-rail-passenger-survey-nrps-spring-2019-main-report/), though apparently not the questions - but I think there isn't much to know, it's "how satisfied are you with" a long list of things. There is a link to more information, but it goes round in a circle; probably something is still on the site but with no link to it (all too common these days).

But really the misleading bit of the article was to call them "new figures released by Great Western Railway...".

Two bits of the methodology to note:

1. Sampling stratified by station footfall. While this is great for ensuring that the results represent the overall passenger base, it does mean that the results tend to represent busy stations (with better facilities and more trains). Perhaps there should be separate stats for "the fringes" of the network.

2. [fieldwork] shifts were aborted if there was only a rail-replacement bus service.

I think I was a participant in this last year. Somebody walking up the train with a bag of questionnaire forms. Handed a form and a pencil. If I remember correctly it asked about your current journey. This was obviously off peak, on SWR, the particular process would not work in peak because the person would not be able to get through the train (certainly so from Reading to London).

That would almost certainly have been an NPRS survey, yes.

The increase in ratings are good news - you can present arguments on their validity and accuracy, but without interviewing every passenger after every journey you could say that about any survey.  The representative sample is diverse enough and the number of surveys large enough to give a good idea of the general feelings of the travelling public.  I know management within GWR take the results very seriously.

Perhaps, just perhaps, it indicates that the average passenger isn't quite as bothered about things such as buffets and padding on seats as some seem to think?  ;)

I think that these surveys are often rigged in order to produce favourable results.

I have taken part in a similar survey on Thameslink.
The surveys are not done on overcrowded trains.
Neither do they include rail replacement buses.
Neither do they include "bad train days" when passengers are advised not to travel.
The questions tend to refer only to the journey being undertaken, with little opportunity to comment on other trips.

Or perhaps it indicates that seat comfort and buffet provision were not included in the survey, perhaps because so doing would result in unfavourable answers.

I would agree if it was a survey undertaken by GWR, but this is undertaken independently by Transport Focus who are on the side of the passenger, not the TOCs, so why would their results be rigged?  I have certainly seen surveys undertaken on a wide variety of trains over the years, including rush hour ones, though whether they get the balance right is always open to examination I suppose.


Seating comfort is a specific question in the survey.  In the case of GWR it was up 4% on Autumn 2018 from a representative sample of 1619.

Catering provision is not a specific question (perhaps it should be?).  It would expect it to have an impact on the 'Overall Satisfaction of the Train' question though if people thought it an important enough aspect of their journey.  The 'Overall Satisfaction of the Train' results were up 7% on Autumn 2018 from a representative sample of 1650.