Archive:Tutorial:Creating a statistical article
It is easy to create a new Statistical article or enrich an existing one by following the step-by-step descriptions below. Statistical articles form the main part of Statistics Explained, presenting and explaining statistical data on a specific topic to the general public. But they also have numerous links to different levels of metadata and to the very latest and most detailed data available, thus making them a user-friendly entry point even for specialists.
Your audienceStatistics Explained is for the general public, not for statisticians and database specialists. While writing, keep in mind a 14-year old lacking prior knowledge but not intelligence. Use simple non-specialist language and do not assume terms and concepts are familiar or even known. This does not mean you must be imprecise or may not use technical terms."Concepts can always be explained by links to Glossary pages (e.g. NUTS). If necessary, create a new Glossary item, using the appropriate and one of the many published glossaries; see 'Creating background pages' (in preparation).Statistics Explained allows anyone within Eurostat, but not outside, to contribute by writing new articles or correcting and updating existing ones. They will become publicly visible after a quick and short validation by D4 Dissemination and by the unit responsible for a statistical theme.Just do it!
If you think something needs to be told or explained, just write it! It can be done very easily and rapidly. By inserting existing material in the Model, it is possible to create a very impressive-looking Statistical article in less than one hour!
This tutorial contains two types of text. The procedures you must follow to create a new article are in plain text. Conventions, rules and recommendations are in a box at the appropriate place.
If you encounter a problem, ask a colleague who already created a Statistical article in Statistics Explained. Or contact the Statistics Explained team (Marc Debusschere, Veronika Lang, Uguzhan Turkoglu, Pavel Borkovec, Ulrich Wieland).
Don't hesitate, you will be surprised how simple solutions usually are! Your questions will improve this Tutorial. Or you can improve it yourself.
Necessary preparation
What to write about
- Take a look at existing Statistics Explained articles, especially in the category or categories your article would fit in.
- Take a look at the Model for a Statistical article.
- Every set of statistical data contains many different 'stories', ways of looking at the data, depending on the aspect or comparison or chronology you want to emphasize. Tell a coherent story which hangs together and is not an enumeration of items. It should not be too short and about one small aspect only, but neither should it be too long or covering absolutely everything in detail for a given statistical theme or subtheme. If it is a general overview it must not be too detailed. If it is detailed, it should cover only a limited area.
- Recycling published and already validated content is a convenient way of getting subject matter. But update it if possible with the latest figures! Publications or parts of publications already containing 'stories' (i.e. some text and analysis) are most suitable. Examples are Statistics in focus, entirely or partly, and chapters from the Yearbook or Regional yearbook. If an article is based upon published tables or graphs only (as in Data in focus or Eurostatistics - Data for short-term economic analysis) you will need to find a text elsewhere or to create it yourself.
- You can of course also write an entirely new story which you think should be told. Explore your data and select what is astonishing or surprising or interesting about them. This can be an evolution over time, a comparison between countries, regions, an overview of a statistical (sub)theme, …
Organizing the material
- Draft a brief outline of the story (Main statistical findings) you want to tell.
- Structure it into introduction, necessary subdivisions, conclusion.
- Collect additional information on 'Data sources and availability', 'Context', 'Further Eurostat information', 'External links'.
The content of chapters
- Data sources and availability: briefly describes where the data were obtained (e.g. a particular survey) and the problems, if any, in obtaining them or in converting them for European use (comparability).
- Context: treats the reasons and uses for the data: the legal basis, the policy context, why society as a whole or groups in society (business, policy makers) need them.
- Further Eurostat information: links to more recent and/or more detailed information available on the Eurostat web site, either downloadable publications, main tables/graphs/maps , database items, dedicated sections on the web site or other information. Do not exaggerate with the number of items, a maximum of the 10 most interesting ones should suffice, less is not a problem.
- External links: should be as article-specific as possible and as deep as possible (not to the home page of ECB, for instance!) to trustworthy (semi-)official instances; try and include some, they are important as service to users and for google ranking of your article!
Entering Statistics Explained
Go to the [Main Page]. At the end of the brief introduction there is a link to two overviews (which can also be found in the 'Navigation' box on the middle left):
- the Statistical_themes page presents a clickable overview of the Statistical themes and subthemes (the nomenclature of statistical products and product groupings for dissemination purposes);
- the Categories page lists the ad hoc categories into which Statistical articles and other pages have been classified, together with the number of articles and/or subcategories in each category. The statistical themes and subthemes are also set up as categories, but in addition other ad hoc categories may be created whenever they might facilitate users' search for a specific page (for more details, see Assigning to category). A category is any reasonable grouping of a sufficient number of articles, which can lead you from one to another.
The 'Special pages' item of the 'toolbox' on the bottom left provide additional access to other overviews: All pages, Popular pages, Recent changes, … Access a specific article by clicking on a page name or, if you know the exact title, by typing it in the 'Search' box on the left top and by clicking 'Go'.
Logging in and entering 'edit' mode
In order to edit pages, you have to log in. Click on 'log in/create account' on the top right and complete the boxes. Before logging in, you may have to create an account.
Starting a new Statistics Explained page
There are several ways to start an article:
- On the Main page or on any page: type the name of the page you want to create in the 'Search' box (left top), click 'GO'; If the page does not yet exist, you get a 'No page title matches' result, with the possibility to click on 'create this page'. Before doing so, check out the 'Page text matches' below, however. A similar page might already exist, which you could maybe update instead of creating a new Statistical article. This list of 'Page text matches' also gives an overview of possible links which you might insert in those other pages, leading towards your Statistical article, once it exists.
- If in an existing page you create a new internal link (see below, 7.1) towards a page which does not yet exist, the linked word(s) appear in red; by clicking on the link you can start creating the page with the title of the word(s) you put an internal link on (See 7.1).
Article names A 'natural' name, describing the content generally, distinctively and briefly, is to be preferred. Take a look at existing articles to get an idea. The titles of publications or chapters are often quite suitable. E.g. 'Minimum wages' (SIF 105/2008). Without a reference to the time period or Europe, European Union or EU, however.
Importing a Model
Go to Model:Statistical article (via the link above or via search 'Template Statistical article' or via 'Main page'/'Special:categories'/'Template'), click on 'edit' (in the menu above the page title) and copy everything except [[<Category:Template>]] at the very end. Paste this into the new page you want to create.
Inserting content
You can of course write an entirely new 'statistical story' about a given data set, filling out the template you imported.
But if you want to convert an existing publication or part of a publication into the StatXP format, all you need to do is to copy the text you want to insert from the original publication (for example the pdf file of the Year Book or any other publication you can download from the website) and paste it in the appropriate place within the template.
For text absolutely no additional formatting is needed. Just check if some slight rephrasing or re-orginazing might not be advisable. Some comments on specific chapters:
- Introduction: there should be a brief and efficient introduction, telling what the Statistical article is about, in simple language, without superfluous non-informative blabla. If the text you copy has no usable introduction, you have to write one yourself, even if it consists only of two or three sentences.
- Data sources and availability: where and how the data were obtained, can usually be found in the methodological notes, as well as problems, if any, with availability (countries or years missing, for example) or comparability issues (e.g. different definitions or data collection methods in different countries).
- Context: Information on legal basis, reasons for the data collection, uses of the data can usually be retrieved from the introduction or the conclusion of a statistical publication or chapter; sometimes they are in the methodological notes.
- If External links or See also or some subdivisions of Futher Eurostat information have no content, those headings must be removed. All other items must have some content.
Inserting links in text
Inserting links is a way to connect your Statistical article internally (within StatXP) to other articles or to the glossary, but also to interesting external information, on the Eurostat site or elsewhere.
Inserting an internal link
A link is 'internal' if it refers to another page within Statistics Explained. The most common case is a link in a Statistical article leading to a page in the background area, usually a Glossary page, briefly and simply explaining a Statistical indicator, Statistical concept, Statistical method, Survey or Data collection or General concept. To insert an internal link, go into 'edit', select the word or words in the text you want to put a link on and click on the 'Internal link' icon Ab, third from the left on top. The selected words are put within [[ ]] (you can of course do this manually too). If the selected words correspond to an existing page, the link is immediately operational. If this is not the case, they appear in red in page view and several possibilities exist:
- You can rewrite the linked words in such a way they do refer to an existing page (e.g. 'EU-27' instead of 'EU 27');
- You can create a new glossary page, if you think the concept needs explaining (e.g. 'Standardised death rate');
- You can redirect from the linked words to an existing page which is synonymous; by creating a 'redirect page' with as only content #redirect [name of the destination page]
E.g.: the content of the page 'EU' consists only of: #redirect [European Union (EU)].
Inserting an external link
A link is 'external' if it refers to a web page outside of SE, either on the Eurostat web site or on other 'external' ones. To insert an external link, go into 'edit', select the word or words in the text you want to put a link on and click on the 'External link' icon (with the globe), fourth from the left on top. As a result, the selected words are put within [ ] (you can of course do this manually too). They have to be edited further so that they include the URL you want to link to: <selected words>. Do not forget to include <http://> ! The selected words should be a user-friendly designation of the target page of the URL, as specifically as possible. Example. World Health Organisation (WHO)
Inserting an Image
Images can be tables, graphs, maps or pictures. Pictures (photographs or drawings) are as a rule not inserted in StatXP articles; only exceptionally, if they present real additional information, not purely as illustration.
Selecting and copying an Image
Select the image you want to insert, e.g. a table or a graph in a pdf-document (without the heading but including foot notes and source at the bottom), a table in an Excel file, etc., and copy it. Leave the pdf-document open, you will still need it to copy & paste the heading.
Transforming and renaming the Image
Open 'Paint', click 'File/'New' and paste the image via 'Edit'/'Paste' (in order to avoid having blank spaces to the right and below your pasted image, you may first have to click 'Image'/'Attributes' and set both 'Width' and 'Height' at '1' = minimal). Now click 'File/Save as':
- Save in: any location where you can easily retrieve the file (e.g. on desktop or in temporary map).
- File name: copy the full name of the original image (e.g.from the pdf document) but without the beginning 'Map 12.1: '; paste this name after 'File name'. You may have to do this in two movements; if the title is in 2 lines, separated by 'return'.
- Save as type: PNG (*.PNG), the last one.
Uploading the Image in Statistics Explained
Click on 'Upload file', third item of 'Toolbox' to the left. Source file name: click on browse and select the file you want to upload, on the location where you have pasted it temporarily. Destination file name: the same file name appears automatically here, but now separated by underscores instead of blanks appears; add at the end of this title the year(s) to which the data refer, separated by a comma and before brackets (if any), in order to make the file unique. (e.g. Taxes_and_social_contributions,_2007_(%_of_GDP). Click 'upload file'. After thus having uploaded the image into StatXP, the temporary PNG file (on the desktop or elsewhere) is no longer needed and may be deleted.
Editing the Image file
Go to the image file (Toolbox item 4 'Special pages'), click on edit Got to 'Template Image', click on edit, copy in into 'edit' of the image fileCopy content
Inserting an Image in a Statistical article
Position of images in a Statistical article Image 1 is normally put at the very first line of the article, before the introduction (so that in the page view it appears to the right of the introduction). Image 2 and all other ones are normally put all together at the beginning of 'Main statistical findings'. If however the resulting page view is not easily readable or not nice (e.g. if Image 1 is quite large), the images can be moved ad hoc to other locations.
In the edit mode, the format of an inserted image consists of
- the name of the Image file;
- formatting parameters;
- the title accompanying the Image
within '[[ ]]' and separated by '|' (without < >, however, added here to leave the code visible).
Example: [[<Image:Top_30_NUTS_2_regions_with_highest_volume_of_air_freight_in_2006.PNG|thumb|right|350px|Table 2: Top 30 NUTS 2 regions with highest volume of air freight in 2006 and index 2003 = 100>]].
The parameters mean that Images normally are presented as a clickable thumb to the right, with a size of 350px; this way, they cover about 40% of a 'normal' page (depending on browser and settings), giving a good balance between text and Images. Replace the file name and the title in the Model, using copy & paste; add 'Graph' or 'Map' or 'Table', accordingly, followed by a serial number.
Adding links to further information
Three cases can be distinguished: linking to an URL outside Statistics Explained, linking to the data tree on the Eurostat web site or linking to other pages within Statistics Explained.
Linking to an URL outside Statistics Explained
There is no difference between linking to another part of the Eurostat web site and to an external web site. Procedures are quite similar, as a result, for the section 'Further Eurostat information', subsections 'Publications' and 'Dedicated section on web site' and for the chapter 'External links' of the 'Model Statistical article'.
In more detail:
- Further Eurostat information/Publications
The template already contains the formatted line (including bullet)
The item in red has to be replaced by the product code which can be found at the back of all publications or on the web site. The item in blue has to be replaced by the precise and full name of the publication, including subtitle if any, edition or series and issue (rules for doing so will be set up). e.g.
If more publications are to be inserted, the line in the template can be copied/pasted and adapted accordingly. It is recommended not to have more than 10 links to publications.
- Further Eurostat information/Dedicated section on web site
The template already contains the formatted line (including bullet)
The item in red has to be replaced by the part of the URL connecting to the dedicated section on the web site. The item in blue has to be replaced by the name of the dedicated section as it appears on the web site. e.g.
Normally only one dedicated section is mentioned in a Statistical article, but if necessary the line in the template can be copied/pasted and adapted.
- External links
The template already contains the formatted line (including bullet)
The item in red has to be replaced by the URL for the deep link into the external web site. The item in blue has to be replaced by a user-friendly designation of the target page of the URL accompanied by the name of the organisation, abbreviated if familiar, between brackets. e.g.
If more external links are to be inserted, the line in the template can be copied/pasted. It is recommended not to have more than 10 external links.
Linking to the data tree
[TO BE COMPLETED] Two methods:
- Linking to the part of the data tree where all data sets can be seen (name of the part of the tree is embedded in the link):
- Linking to a part of data tree where only subfolder but no data sets can be seen (the link and the reference 'see: Name of subfolder' are provided separately:
- [1] see:
- Government statistics
Insert second part of the URL by right clicking on "+/-" sign on the branch of Data Navigation Tree you want to display, copy shortcut: part after "&open="(e.g. "/popul/livcon/ilc&language=en&product=EU_population_social_conditions&root=EU_population_social_conditions") and paste it into template.
- Put title(s) after "see:" by copy/paste title(s) of Data Navigation Tree subdivision(s) where the tables, graphs and maps can be found (e.g. "Income and living conditions"
Linking to other pages in Statistics Explained
As an extra service to users (and to enhance the 'linkedness' and thus the google ranking) links can be provided to a limited number (not more than 10) of high-quality links to trustworthy (semi-)official external sites (e.g. WHO, ILO, FAO, ECB, UNECE, OECD or NSIs). The links should be as specific to the subject treated as possible and they should be deep links directly to the interesting information, not to the home page!
Assigning to category
At the bottom of the template, replace <Category name(s)> (including < >!) in [[Category:<Category name(s)>]] with the appropriate theme or subtheme (see List of statistical themes and subthemes for the exact names and hierarchy). If you want to have more categories, copy and paste [[Category:<Category name(s)>]] it after the first one, separated by a blank, and change it accordingly (see List of current categories for an overview).
Example for Statistical article 'Transport infrastructure' (without < >):
[[Category:<Transport>]] [[Category:<Regions>]]
Validation of an article
You don't need to do anything for this. The wiki system is to be set up so that validators are notified automatically of any new article or of any change in an existing one. They can compare very easily old and new versions and evaluate the quality of the changes.
Governance rules foresee a quick validation by D4 Dissemination, accepting and making public small changes immediately and sending significant content changes not from the unit owning a page to the unit owning it for validation. All Statistical articles are the responsibility of one unit. Validation should be rapid.