Here’s What Phrases Like “very Likely” and “probably Not” Mean to Different People.
How likely is “likely” for you? Redditor zonination asked users of subreddit / r / samplesize to assign percentages to a number of common likelihood phrases and displayed the results . The sample size is small (46 respondents in total), but the data still looks grouped enough to be useful, plus it is broadly in line with the survey that inspired 23 NATO officers . Two years after their initial study, the zoning changed the data plotted as a great ” schedule of joy ” ( a la Division Joy ).
The graph above is smoothed; you can see a more accurate and unstable graph as well as raw data on the project’s GitHub page . There are a couple of silly outliers (one person is labeled “probably not” as 100%), but the zoning says all outliers are from the same respondent (so they probably should have been eliminated).
There is also evidence of what numbers such as “a few” or “tens” mean to people, but the graph is on a logarithmic scale, which makes it difficult to interpret some of the answers. Of course, the phrases below were devoid of context, which could have completely different meanings. “Some Americans have two houses” obviously means something other than “I left the dishes in the sink.”
The main takeaway is that people interpret the same phrase in different ways , so when you need to communicate an important probability or amount, it is better to use a specific, quantitative language, ideally a valid numeric range.
For more plots,look for #joyplot on Twitter .
Perception | GitHub via Randy Olson