A few interesting things to start your week
Survivorship bias and selection effects, a great resource for selecting HR software, and bottlenecks to evidence adoption
The bread and butter of this blog is long-form articles, and I’m unwilling (or unable) to compromise on quality to churn out more frequent posts. That said, a nice compromise might be to link to some interesting things I’m reading every now and then. Let me know if you like this format!
Ethan Mollick on survivorship bias:
Fascinating and very relevant to personnel selection (i.e. range restriction) but also from a behavioral econ perspective, people’s beliefs seem to anchor on anecdotes (“success narratives” in this case), which are subject to selection effects, including survivorship bias. Anecdotes, stories, and “case studies” are overrated.
Speaking of selection effects, I highly recommend this NYT podcast on how the science of alcohol’s impact on health has developed over time. In summary, older studies comparing “non-drinkers” to “light/moderate drinkers” failed to account for the fact that many “non-drinkers” were former drinkers who quit because they developed health problems, perhaps due to alcohol!
One of my least favorite aspects of working in HR is having to deal with HR technology vendors. The mis-aligned incentives of the sales process mean that the software you see in the demo hardly ever lives up to reality, and the quality of service you get after the sale hardly ever lives up to what you got before you bought.
That’s why I was really interested to come across this site: SelectSoftware Reviews - an independent service that reviews HR software, and will even work with you 1-1 to help you make the best decision.
You can check them out, and even book a free consultation using this link to get some expert help to save you and your organization a lot of pain!
What are the bottlenecks to evidence adoption? This study looked at this question in public policy and found that a leading explanation was “organizational inertia”. A takeaway for People Analytics pros: when attempting to influence behavior with data, utilize existing communications / communication channels where possible.
