![]() | This is an essay. It contains the advice or opinions of one or more Wikipedia contributors. This page is not an encyclopedia article, nor is it one of Wikipedia's policies or guidelines, as it has not been thoroughly vetted by the community. Some essays represent widespread norms; others only represent minority viewpoints. |
Statistics are available for the pageviews, or traffic, that articles receive. For a broad introduction, see Wikipedia:Pageview statistics. The focus here will be on applications of these statistics in determining a primary topic.
Several important tools are:
- Pageviews: It shows the pageviews for an exact title, whether it is an article or a redirect. It displays daily or monthly views for the selected time period, and it allows for comparing the views for different titles.
- Redirect views: It shows the total views of a single article for a given period, with the number broken down by views for the exact title and for each incoming redirect. This can be helpful for gauging the overall popularity of articles.
- Massviews: It shows the pageviews of each page linked from the specified page. Primarily useful for gauging the relative popularity of articles linked from a disambiguation page. (Note that you'll get some irrelevant links like Help:Disambiguation which need to be filtered out manually.)
Factors to bear in mind when analyzing pageviews include:
- Time period selected. Typically, the longer the time period the more robust the conclusions, but circumstances may call for a shorter period, or a comparison of two or more periods.
- Did all articles being compared exist for the entire period? Did all articles being compared have the same title for the entire period? If any article(s) were moved during the period, be sure to include all titles of all articles, or adjust the chosen period accordingly. Unless using the Redirectviews tool, comparisons might not be representative if traffic from incoming redirects is not included.
- Are there any spikes during the period? If so, consider whether or not to discount one time events that could skew long-term trends. Events which can cause spikes include being in the news, page creation, and opening a Wikipedia discussion. Are pageviews fairly consistent for all articles during the period? If there are large variations, consider what might be causing them. If any articles are fairly new, consider whether the trends shown are likely to continue into the future. Are the total views for all compared articles very low? Very low numbers may not be statistically meaningful; differences in linking on Wikipedia can also have a disproportionate effect.
- How many articles are being compared? If only two or three, consider whether WP:ONEOTHER and WP:TWODABS are relevant.
Two major aspects of a primary topic are long-term significance and usage. Pageviews have some, indirect, relevance for both: sustained high views over many years could be a result of greater long-term significance, and big differences in views may also indicate differences in usage. However, pageviews contain no direct information about what readers search for. One reason is that they show what readers get, not what they look for. For example, if one of the articles is at the base title, then some proportion of the traffic it receives will come from readers looking for other topics with the name. Another reason is that a substantial proportion of views come from readers who arrive by following internal or external links. Typically, traffic coming from such links is much greater than traffic coming from readers searching directly from within Wikipedia. The overall traffic is dependent on results from Google and other external search engines, as well as how well integrated an article is with the rest of the encyclopedia and how popular the pages are that link to it. This has the following implications:
- Pageviews for articles (which have incoming links) are not directly comparable to pageviews for disambiguation pages (which typically do not have incoming links (and should not - direct links to DAB pages are errors per WP:INTDAB, and industrious WikiGnomes labor to fix them)). In a hypothetical scenario of absence of traffic from incoming links, if an article located at the primary title An example received twice as many views as An example (disambiguation) (which is linked from a hatnote at An example), this would entail that half of readers who load the article proceed to click through to the dab page, so the article would only be the topic sought by half of them. Given that a lot of the traffic for an article comes from links (which are presumed to lead readers where they intend to go), that article will still receive much higher views than the disambiguation page even if it is not the primary topic with respect to usage.
- Similarly, pageviews for articles are not comparable to pageviews for redirects. For example, the article London, Ontario receives a lot more hits than the redirect London, England, but it obviously isn't the primary topic for London.
Care should be taken when evaluating terms that are not identical to those of the articles concerned. For example, in deciding the primary topic for "MSG", the choice might be between Monosodium glutamate and Madison Square Garden. Although those two articles receive a similar number of views, Monosodium glutamate may be referred to primarily as MSG, while Madison Square Garden is usually referred to by its full name and is much less frequently known as "MSG". The relevant question when looking at usage in these type of cases is not which of the articles is more popular, but which one is more likely to be sought by readers searching for "MSG". If there is a primary topic for "MSG" it can become a "primary redirect"; otherwise it can be a dab page; but pageview statistics might not be helpful in making that decision. Readership usage in such cases can be inferred from the Clickstream data (or from its visual frontend the Wikinav).