Is the thumbs up (👍) emoji passive aggressive?
Posted: Sun Jun 21, 2026 6:37 am
SURVEY TOOLS USED
I posted this question as a poll on instagram.
49% of my followers responded that they think reacting with
is passive aggressive.
n=104 (international, mixed gender, ages ~26 to ~35)
Doing it again
Then I had a chance to ask the question again in a Customer and Market Insights class as part of a demo the professor was doing about survey platforms. The question was phrased slightly differently but I don't think that explains the difference in results (my first dataset was obviously skewed, though it's probably more useful to me as it is more representative of my own environment).
Imagine a friend reacts to one of your messages with a thumbs up emoji (). Would you consider this to be "passive aggressive"?
38% say yes, 62% say no. The same number said yes it if was a colleague.[0]
When you message with friends, do you use the thumbs up emoji ()?
76% say yes. 24% say no. The same number said yes for colleagues.[0]
n=100 (UK based, mixed gender, mixed ages)
on the reliability of the results
The replication attempt clearly shows inconclusive results, but I feel like you could find repeatable results if you segmented the population.[1]
why does this matter / what might these results imply
I wrote about this here: https://phone-less.com/49-of-my-instagr ... aggressive
[0] I don't think that the friends/colleagues distinction here is accurate; the questions were given in the same survey, there probably wasn't enough separation between them. There may be some element of people just trying to get the survey done fast (they get paid to do it.)
[1] I wonder what replicability means; if something is replicable but just with a specific group of people, or a specific person, does that make a result valid / replicable / scientific? That is probably useful information about that person that holds consistently over time (maybe just in a specific context). Does science always have to be generalisable?
- Instagram polls
- Prolific: https://www.prolific.com/ - lets you conduct surveys with real people really fast. It's not free (100 UK participants for a 1 minute survey was just under £40.00. It took roughly 10 minutes for 100 people to take it, from the time it went live.)
I posted this question as a poll on instagram.
49% of my followers responded that they think reacting with
is passive aggressive.n=104 (international, mixed gender, ages ~26 to ~35)
Doing it again
Then I had a chance to ask the question again in a Customer and Market Insights class as part of a demo the professor was doing about survey platforms. The question was phrased slightly differently but I don't think that explains the difference in results (my first dataset was obviously skewed, though it's probably more useful to me as it is more representative of my own environment).
Imagine a friend reacts to one of your messages with a thumbs up emoji (
). Would you consider this to be "passive aggressive"?38% say yes, 62% say no. The same number said yes it if was a colleague.[0]
When you message with friends, do you use the thumbs up emoji (
)?76% say yes. 24% say no. The same number said yes for colleagues.[0]
n=100 (UK based, mixed gender, mixed ages)
on the reliability of the results
The replication attempt clearly shows inconclusive results, but I feel like you could find repeatable results if you segmented the population.[1]
why does this matter / what might these results imply
I wrote about this here: https://phone-less.com/49-of-my-instagr ... aggressive
[0] I don't think that the friends/colleagues distinction here is accurate; the questions were given in the same survey, there probably wasn't enough separation between them. There may be some element of people just trying to get the survey done fast (they get paid to do it.)
[1] I wonder what replicability means; if something is replicable but just with a specific group of people, or a specific person, does that make a result valid / replicable / scientific? That is probably useful information about that person that holds consistently over time (maybe just in a specific context). Does science always have to be generalisable?