Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Wikipedia Businesses

Wikipedia Edits Have Massive Impact on Tourism, Say Economists (theguardian.com) 49

Forget glossy travel brochures and whizzy online sites; one of the most cost-effective ways tourism chiefs can drive business to their towns or cities is by updating their Wikipedia page. From a report: An experiment by economists at the Collegio Carlo Alberto in Turin, Italy, and ZEW in Mannheim, Germany, found that a few simple edits to a Wikipedia page could lead to an extra $130,000 a year in tourism revenue for a small city, underscoring the power of the free online encyclopaedia. The researchers randomly selected cities across Spain to receive targeted improvements to their Wikipedia pages, adding a few paragraphs of information on their history and local attractions, as well as high-quality photos of the local area.

It didn't take an expert, either. Most of the content added was simply translated over from the Spanish Wikipedia into either French, German, Italian or Dutch. Doing so had an immediate and remarkable effect: adding just two paragraphs of text and a single photo to the article increased the number of nights spent in the city by about 9% during the tourist season. In some instances, the increase was even larger. For cities with barely anything on their Wikipedia pages, a minor edit could raise visits by a third.

This discussion has been archived. No new comments can be posted.

Wikipedia Edits Have Massive Impact on Tourism, Say Economists

Comments Filter:
  • Scotland (Score:4, Funny)

    by phalse phace ( 454635 ) on Friday September 18, 2020 @12:46PM (#60519376)

    Wikipedia Edits Have Massive Impact on Tourism, Say Economists

    How's Scotland's tourism industry doing?

    Most of Scottish Wikipedia Written By American in Mangled English [slashdot.org]

    • Wikipedia Edits Have Massive Impact on Tourism, Say Economists

      How's Scotland's tourism industry doing?

      Not to be stereotypical, but aren't they too drunk to care?

  • by Anonymous Coward

    So, an increase in tourism followed modifying Wikipedia articles.

    But, to assume the Wikipedia article is the *reason* behind the increase is somewhat ridiculous unless you can completely rule out all other variables.

    Did the economy do well the year tourism increased? The increase bay be due primarily to an increase in disposable income.

    • It could be the increase in tourism that caused the edits to the related Wikipedia pages. People added things that they thought were missing but would have been useful to know.

      • It could be the increase in tourism that caused the edits to the related Wikipedia pages.

        Nope.
        The summary and the news article both mention that the edits were done by the researcher themselves.
        (And were banned by the dutch wikipedia admin for being too commercial).

        The increase was measured after the edits.
        But the news article fails to mention what the researcher used to establish a baseline against which to measure this increase.
        (Did they also did a random selection of cities that they did not edit as a negative control ?) or other form of analysis.
        (Did the cities' tourism office report an inc

    • by ranton ( 36917 ) on Friday September 18, 2020 @01:18PM (#60519510)

      to assume the Wikipedia article is the *reason* behind the increase is somewhat ridiculous unless you can completely rule out all other variables. Did the economy do well the year tourism increased? The increase bay be due primarily to an increase in disposable income.

      For god's sake, how do you think any research is done? Just use a control group. Very little of their methodology was mentioned in the article so who knows if it was worthwhile research, but it wouldn't be too hard to design a study to mitigate for the factors you mention. At a high level just pick a good distribution of 100 cities you will make the Wiki updates to and 100 cities you will not. Then check the change in tourism in each city. Perhaps monitor other Wiki changes not done by your research team to see if those cities should be removed from the study.

      That is just the basics. There are plenty of techniques they could use to ensure they aren't just measuring normal economic improvements, among other factors.

    • by Anubis IV ( 1279820 ) on Friday September 18, 2020 @01:24PM (#60519538)

      So, an increase in tourism followed modifying Wikipedia articles.

      But, to assume the Wikipedia article is the *reason* behind the increase is somewhat ridiculous unless you can completely rule out all other variables.

      Did the economy do well the year tourism increased? The increase bay be due primarily to an increase in disposable income.

      They asserted that there was a causal impact in the opening line of the paper. [hinnosaar.net]

      It's not difficult to control for such variables, which they did. The 9% improvement the summary talked about? That's 9% over the control group according to the paper.

      They selected a randomized set of cities across Spain, updated those pages on Wikipedia across multiple languages, and then waited to see what the results were. Simultaneously, they randomly selected a different set of cities for their control group, then used those cities as a basis for comparison, allowing them to control for whether changes in local or national economy, overall tourism trends towards Spain, etc., etc. etc. had made a difference. And one of the benefits of using Spanish cities is that the hotels keep a record of the nationality of the tourists staying overnight, providing the scientists with an additional indication of whether their changes to Wikipedia had been effective (i.e. if a city sees a tourism boom in general, it proves nothing, but if it sees a boom specifically from the people who speak the languages of the pages they had edited, that would point towards a causal relationship, which in a quick skim of the paper appears to be what they saw). Oh, and they also performed a statistical analysis to make sure that their selection of control cities did not affect their results.

      I love how the first thing that everyone on Slashdot does is assume that no one who literally does this job for their living knows how to properly control for variables. While there are some scientists who cut corners, most of the time they've done their homework, especially when the research took over a year to perform in total.

      • by penandpaper ( 2463226 ) on Friday September 18, 2020 @02:37PM (#60519868) Journal

        >knows how to properly control for variables.

        Well if you have seen some of the code I have seen, programmers don't know how to properly control their own variables. Maybe it's projection.

        • by dgatwood ( 11270 )

          No matter how good you get at programming, you never outgrow the tendency to accidentally forget to increment/accumulate at the end of a sufficiently complex loop.

          Maybe they should repeat this experiment in Russia. After all, <Yakov surname="Smirnoff">in Soviet Russia, variables control you.</Yakov> :-D

        • Well if you have seen some of the code I have seen, programmers don't know how to properly control their own variables. Maybe it's projection.

          I just assign one variable to be another, and that other back to the first and then sell the software as "Powered by AI".

      • by ljw1004 ( 764174 )

        I love how the first thing that everyone on Slashdot does is assume that no one who literally does this job for their living knows how to properly control for variables. While there are some scientists who cut corners, most of the time they've done their homework, especially when the research took over a year to perform in total.

        Thank you for your summary of the paper.

        We all know that correlation isn't causation. But I think that correlation is *correlated* with causation... (at least, scientific reports of correlation are correlated with scientific discoveries of causation).

        • The "correlation isn't causation" is about observational studies.

          They actually did a randomized experiment, so that makes a case for causation (not an airtight one, but this kind of study rarely if ever is).

  • by xack ( 5304745 ) on Friday September 18, 2020 @01:17PM (#60519502)
    WIkipedia has a sister project called Wikivoyage that is specifically designed for tourism info. I think the Wikimedia foundation should promote that to tourists more.
    • Personally, I look up information on a city for a random reason, and it either ends up in the bin, or as a place where maybe I want to go at some point in the future for whatever reason. For people like me, having interesting content of varied nature (not just tourist attractions) can leave a good impression. It doesn’t even need to be cities that are considered tourist destinations— knowing that The Dish Room in Burlington, CO is owned and operated by a local guy that went to culinary school i
      • I use Atlas Obscura for this sort of thing, and while it's not all that often that I find something that appeals to my interests and isn't far out of the way, I find enough to make it worth checking out my planned routes to see if they mention anything along the line.
  • Study limitations (Score:5, Informative)

    by skoskav ( 1551805 ) on Friday September 18, 2020 @01:22PM (#60519524)

    Link to actual study: https://papers.ssrn.com/sol3/p... [ssrn.com]

    Some of my quick takeaways of the possible limitations of the study:

    1) They acknowledge that a significant part of the effect could just be tourism "stealing" from neighboring cities.

    2) The effect was more significant if the original article was nearly empty.

    3) They used control groups in a perhaps unintuitive way:

    To ensure balance in the treatment and control groups, we used a stratified randomization design. We ordered the 60 cities by the total number of tourists. Then we divided the cities into ten groups of six cities each. Within each group, we randomly assigned the city to one of six treatments. The six treatments were as follows: treat the city page in one of the six possible language pairs (Dutch & German; Dutch & French; Dutch & Italian; German & French; German & Italian; French & Italian). Hence, 120 city pages were treated and 120 pages remained as controls.

    It's not an issue per se to group the different wiki languages like that, but I did not find a clear justification for it. If not properly accounted for it could be an indicator of p-hacking, by looking at their collected data whichever way is the most favorable for a positive outcome, and downplaying the remaining data.

    4) The Dutch wiki admins were assholes.

    • by Anubis IV ( 1279820 ) on Friday September 18, 2020 @01:31PM (#60519568)

      4) The Dutch wiki admins were assholes.

      The word "Dutch" was unnecessary in that sentence.

    • a significant part of the effect could just be tourism "stealing" from neighboring cities.

      Perhaps. But if I run a restaurant in town X, I don't care if my competitors in neighboring town Y get less business.

      • Yeah, the local businesses wouldn't care whether the tourists are just diverted or if the country's tourism sector is actually affected. But it would matter for the government.

        So the study's design limited the effect from being better quantified and understood.

    • by pr100 ( 653298 )

      It could be p-hacking if they didn't determine that this was going to be their methodology in advance.

      This is pretty much the same in every study: if you gather your data and then decide how to slice and dice it, then you're in questionable territory. You need to decide how you're going to do the analysis *before* you have the data.

  • a few simple edits to a Wikipedia page could lead to an extra $130,000 a year in tourism revenue for a small city,

    Does this just divert visitors away from other nearby locations or does it actually generate new business and attract visitors to make additional journeys.

    • by b0bby ( 201198 )

      From the point of view of any particular city, that may not matter.

    • Tourism has plenty of zero-sum scenarios, but the more places people want to go, the more time they will spend traveling.
    • Every second of every day is a zero sum situation. So probably. But it might increase the chances of going to a region if there is something nearby. E.g. my wife wanted to go to Sevilla in Spain but we had to stop in Barcelona and Madrid for the airports so we spent time in each of those locations. And we also read about towns that were a train ride away from Sevilla.

      So yes that probably took away time we could have used to go to the Netherlands or Austria but usually people journey a bit around the d

  • Those live cams seem to quite effective as well.

  • Tourism industries have a knack for covering up all of the bad stuff that happens in those areas.

    If I go to a resort and there is a high risk of me get robbed and assaulted the moment I step off of the grounds of the resort, I'm not going to want to go to that place either.

Technology is dominated by those who manage what they do not understand.

Working...