Issue link: https://resource.opensesame.com/i/1474251
Caveats The L&D Global Sentiment Survey is a sense check on the sentiment of people working in Learning and Develop- ment (L&D). For this reason, it is designed around a single question, the same each year, "What will be hot in workplace L&D in [year]?" with the aim that people can respond quickly to the survey. This design – of an anon- ymous, online poll – and the methods used to canvas votes, mean some caveats must be applied to the results: Respondents are likely to be tech-savvy Most respondents are invited to participate via social media and email. They are, therefore, a self-selecting group. Because they are contacted – and answer – electronically, respondents are certainly users of technology, and probably more likely to feel positive- ly about technology than the general population. This method of canvassing votes means anyone working offline is excluded. Respondents are largely unqualified We do not know for certain that the respondents work in L&D. Some are approached via direct messaging on LinkedIn because of their job title, but could have moved to non-L&D roles. People approached via email will have shown some interest in L&D in the past, but may no longer. We cannot guarantee that any respondents worked in L&D when voting, or that they have not passed the voting link on to others unconnected to the field. We cannot control who responds to links shared on social media. However, completing the survey would be a dull pastime for anyone not in our field, and we think it unlikely the results are badly affected by non-quali- fied respondents. Year-on-year comparisons may be invalid Because the survey is anonymous, it is impossible to guarantee that the same people vote each year. In fact, as the numbers on the survey increase each year, it is certain they are not. This could lead to variations between surveys arising from changes in the surveyed population, not in changes to senti- ment of the originally surveyed population. Understanding of the options may vary To make the survey quick to complete, no defini- tions are provided. If they were provided, this would give an illusion of certainty – we would have no guarantee respondents would actually use them. We have to accept that not all respondents will agree on what the options mean, and while not all fluent English speakers will agree non-fluent English speakers may vary even more widely in their under- standing. Furthermore, there are some terms – notably Curation – that are not part of regularly-used English, so some non-native speakers may disregard them in voting. (As far as we can tell this has not happened for Curation.) Voter selection may skew results In some countries, respondents are largely attracted to the survey by individuals or organisations promi- nent in that country. In Sweden, for example, the survey was mostly promoted by a single company, and in Ireland mostly by a single institute. Potentially, this could skew results. Social media could influence voting It is quite possible that proponents of a particular technology or methodology could campaign for it and boost its standing in the survey. There have been no signs of this happening (eg sudden spikes in support for a particular option). Less obviously, it is possible that prominent influencers might inadvert- ently influence the survey (eg by making public predictions) with the same results, but we have also seen no evidence of this. Donald H Taylor, Global Sentiment Survey 2022 20