22

As promised here I've done an analysis of questions to date to help understand what we've deemed on-topic so far.

From the beginning of the Beta to circa 1500 GMT on 14th November, 261 questions had been asked, of which 14 (5.3%) were closed. Although only two of these questions were closed as off-topic, I’ve disregarded all closed questions in the analysis that follows.

Breakdown by Topic Area

I’ve categorised the questions into 6 “Topic Areas”; the sort of questions I’ve allocated to each are shown at the end of this post. For some questions, a case could be made for it to sit in more than one topic area, but I’ve made my best judgement and I don’t think it materially affects the analysis.

Pie chart of questions by topic area

Over half our questions have been asked about Locating Records (30%) and Best Practice/ Research process and methodology (28%). Researching specific ancestors (14%), Interpreting records (13%) and Applying technology (12%) make up almost 40% between them. My view is:

  • that this is a healthy mix of on-topic subject matter
  • I’m encouraged that we’ve closed so few questions as off-topic
  • we should expect the proportion of Best practice questions to reduce as the ‘obvious’ topics are covered and it becomes harder to avoid duplicates
  • we should expect the proportion of questions on Locating records, Interpreting records, Researching ancestors and Applying technology to increase. There’s a potentially limitless supply of these out there; our challenge will be to get them asked here rather than somewhere else, and get them asked well.

Analysis of Votes by Topic Area

For each topic-area I’ve extracted the minimum, maximum and average votes per question. This should be a reflection of how good the questions are viewed to be, but as a community we have not yet thrashed out our definition(s) of good, and whether voting means “I like this question”, “I’m interested in the answer to this question” or “I think this is a good question”.

Votes by topic

On average, we have a preference for questions on Technology, Best practice and Interpreting records. Might that be because these are the more generally interesting topics?

Minimum votes in all the topic areas are pretty consistent, proving nothing except there are always some questions that are so narrowly focused or so badly written they don’t appeal.

Maximum votes in most topic areas are similarly consistent. However, Best practice questions have the widest range of popularity (even discounting the single question that scored 24 votes). And Specialist questions don’t get as much love as the others.

Overall Distribution of Votes

I was curious to see what the overall voting pattern looked like.

Distribution of votes by number of questions

1520 votes had been cast in total at the time the snapshot was taken. The average question gets 6 votes. 9.7% of the questions were average by this measure, and 52.6% of the questions were below average, making 62.3% in total average or below.

Description of Topic Areas

This is an overview of the sort of questions I’ve allocated to each topic area. If it’s generally agreed that this is a useful breakdown, I’m willing to work on it some more in discussion with others) to form the basis of guidance on what is on- and off-topic here, along the lines of the post Luke pointed to at https://english.meta.stackexchange.com/questions/273/what-questions-are-on-topic-and-off-topic-here.

Application of Technology

  • Use of specific software products and web sites/technologies
  • How do I do X with package/website Y
  • Establishing criteria for choosing software products
  • Data exchange formats and other technical standards (GEDcom, FHISO etc.)
  • DNA testing
  • Use of GIS technology (Geographic Information Systems)

Best practice/research process and methodology

  • How best to document and present what I’ve learned
  • Ethics and legalities (copyright, privacy, sensitivity)
  • Numbering systems
  • Working with personal or place names
  • Working with sources

    • Citation
    • Evaluation
  • Genealogical Proof Standard and other best practice guidelines
  • Organising and tracking research activities and results
  • Approaches to research

    • Getting started
    • Working with others
    • Inferential genealogy
    • Person based versus record based research
  • Document and artefact handling and preservation

Locating Records

  • How can I find records [about X event] [in P place] [around time T]

  • How can I find a specific record/document

  • How can I find information about a specific topic

Interpreting Records

  • Reading and comparing handwriting

  • Working with records in foreign languages

  • Dating, identifying and extracting information from photographs

  • Understanding terminology in records

  • Understanding record types and their contents

  • Using specific record types in research

  • Evaluating the reliability of specific sources

Researching Specific Ancestors

  • How do I learn Y fact about a specific ancestor

  • How do I learn about the historical context of a specific ancestor

  • How do I learn X historical fact, knowing which may lead me to more information on an ancestor

Other Related Specialist Topics

  • Heraldry
  • Origin and meaning of surnames
  • Government policies relevant to genealogists
  • Academic and other studies of genealogists and their networks
9
  • 7
    You've done a superb analysis! I'm very encouraged by where you've placed us. I think your "topic areas" should be considered as content for our FAQ.
    – lkessler
    Nov 20, 2012 at 18:56
  • 2
    +1; this is really good stuff.
    – GeneJ
    Nov 20, 2012 at 20:34
  • 2
    Thanks for doing this. It is quite helpful. +1
    – Luke_0
    Nov 21, 2012 at 0:38
  • Thank you for a most excellent analysis of the question pool to date. I am wondering how much you could rely on the question tagging to determine the top-level "area" to which a question belongs? If that were true then some programmer-type among us could automate the kind of data mining and charting you've done here. That way the computers can do what they do well (show you the interesting data), and the humans can do what we do well (interpret the interesting data.)
    – fbrereto
    Nov 22, 2012 at 0:45
  • @fbrereto, if every question was tagged with one of the topic areas, then that would work well. There's no one-to-one correspondence between any existing tags and the topic areas however.
    – user104
    Nov 22, 2012 at 9:41
  • excellent work! Did you pull the data by hand or did you use the stackexchange api?
    – Duncan
    Nov 26, 2012 at 13:23
  • @Duncan, I screen-scraped the data using outwit.com/products/hub -- easiest for me because I knew how to use it without a learning curve. I'll bet there's a better way though.
    – user104
    Nov 26, 2012 at 13:30
  • 1
    I think you should move the 'description of topic areas' into an answer. Then the discussion could be directed to that, AND we could upvote it as a good answer. This seems like too much good work to have as an unanswered question.
    – Duncan
    Nov 27, 2012 at 12:28
  • @Duncan, I'm planning to move the description of topic areas into a set of new questions, one per topic area to drive the on- and off-topic discussion as I suggested.
    – user104
    Dec 3, 2012 at 13:09

0

You must log in to answer this question.