The data in this survey are produced from a combination of vote analysis and examination of opportunity data. For example, the number of opportunities with more than six essential skills can easily be counted.
The data based on votes is calculated as follows: for each opportunity, we look at all the votes for the given question and present the proportion of opportunities for which over 50% of the votes were either "bad" or "very bad".
This work was not completed by statisticians. We've squeezed this in alongside paid client work, which has limited our ability to be thorough. It is not rigorous. Its intent is to give a general sense of content quality on the marketplace. We hope that it will be useful enough to make the case that more substantial research is needed.
- Individual respondents are not tracked. For the purposes of this page of results, each vote per question is assumed to be a different person
- There is no mechanism to prevent multiple votes on a single question from a single individual
- There is no selection of respondents: anyone is free to participate. Due to the subject matter we suspect the majority of participants will have experience using the Marketplace, and an active interest in it, but we have no mechanism to validate this
- There is no mechanism to prevent mass voting (though we have no indication so far that this has occurred)