Hi Nathan, I’m one of the co-organizers along with Juliana and I’ve thought a lot about quantitatively measuring attrition rates and the types of people more interested in EA.
We found it hard—at least on the level of one club—to measure things like attrition rate for a few different reasons:
First, there are so many factors that may cause someone to attrit from a club, ranging from them not being a good fit to them simply not having the time (as @DavidNash mentioned).
These factors also are so variable: as Juliana mentioned, if she didn’t like interacting with her group (me especially) as much, she probably wouldn’t have continued engaging in EA. One bad session or rude person in your group may make the difference between a person continuing to engage or leaving.
These factors all also have such small effect sizes and also intersect with each other, to the point that it becomes very hard to suss out any sort of causality or clear picture from quantitative data.
This is where (in my experience) these more qualitative approaches become more useful, as we are naturally attuned to gathering peoples’ impressions.
Second, there’s going to be a lot of selection bias: most people (at least I would suspect) in most universities aren’t interested in EA (again, for many factors) and so any confusion matrix of people ‘interested/not-interested in EA’ and ‘whether they join’ is going to be very heavily weighted towards ‘not-interested x not join’. And so the vast majority of people who come to an information night or express some sort of interest are not leaving for any particular reason, but rather because they just aren’t interested in EA [enough].
But as you mentioned, there may be a lot of value in sort-of qualitatively-quantitatively measuring attrition rates on the scale of CEA—instead of trying to find reasons as for why people are not staying on a group-level, the CEA Groups team could survey reasons for why group organizers think people leave, and perhaps use that to create helpful resources.
Solid piece. I like lists of things and I appreciate you taking the time to write one.
I sometimes wonder how to combine many qualitative impressions like this into a more robust picture. Some thoughts:
Someone could survey groups on attrition rates
Someone could ask how many people group leaders recalled who were in each group type
Hi Nathan, I’m one of the co-organizers along with Juliana and I’ve thought a lot about quantitatively measuring attrition rates and the types of people more interested in EA.
We found it hard—at least on the level of one club—to measure things like attrition rate for a few different reasons:
First, there are so many factors that may cause someone to attrit from a club, ranging from them not being a good fit to them simply not having the time (as @DavidNash mentioned).
These factors also are so variable: as Juliana mentioned, if she didn’t like interacting with her group (me especially) as much, she probably wouldn’t have continued engaging in EA. One bad session or rude person in your group may make the difference between a person continuing to engage or leaving.
These factors all also have such small effect sizes and also intersect with each other, to the point that it becomes very hard to suss out any sort of causality or clear picture from quantitative data.
This is where (in my experience) these more qualitative approaches become more useful, as we are naturally attuned to gathering peoples’ impressions.
Second, there’s going to be a lot of selection bias: most people (at least I would suspect) in most universities aren’t interested in EA (again, for many factors) and so any confusion matrix of people ‘interested/not-interested in EA’ and ‘whether they join’ is going to be very heavily weighted towards ‘not-interested x not join’. And so the vast majority of people who come to an information night or express some sort of interest are not leaving for any particular reason, but rather because they just aren’t interested in EA [enough].
But as you mentioned, there may be a lot of value in sort-of qualitatively-quantitatively measuring attrition rates on the scale of CEA—instead of trying to find reasons as for why people are not staying on a group-level, the CEA Groups team could survey reasons for why group organizers think people leave, and perhaps use that to create helpful resources.