You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more
New research shows that the fear of smart phones and social media was built on a castle made of sand. Turns out almost all of previous research never bothered to validate their assessments of smart phone use - and that appears to have been a HUGE mistake. (.....)pic.twitter.com/1ETRiuRM9J
.... Almost all research examining the harmful effects of social media (SM) have used self-reports of SM. But, new research has found, such self-reports are hardly related to actual SM use. In fact, the average correlation between self and actual is about .20 .....
....What does a correlation of .20 mean. No researcher would ever agree that a correlation of .20 is ok if two assessments are supposed to be measuring the same thing. Another way to think of it is that self-report and actual SM use share only 4% of their variance.....
Or a way to visualize is with this scatter plot. Regardless of your understanding of this plot – that is a lot of random noise going on there! If you scored high on self-report of SM it tells you almost nothing about your actual SM use.pic.twitter.com/1DnPXSnm60
In short–all research on SM that used self-report SM is not examining actual SM use. What are they examining? Who knows! But think of all the articles, talks, books, meetings,etc. that have all taken place. All based on research that didn’t validate their primary variable.
Research here https://psyarxiv.com/6fjr7 by @davidaellis @BritDavidson @H_Shawberry and Kristoffer Geyer
How did previous scholars justify the use of self-report of SM? Turns out they often cited incorrectly. EG This study cites a validation indicating that the validation study found that self-reports were "consistent" with actual behavior....pic.twitter.com/SAxc2Xx39Q
....But the validation study the city did NOT find this. In fact the validation study found similar low relations between self-reports and behavior for everything except for scheduled activities (like having a job).pic.twitter.com/0JJxoZNeWY
See also “The Accuracy of Self-Reported Internet Use—A Validation Study Using Client Log Data” by @mscharkow, which finds pretty much the same for self-reported Internet usehttps://doi.org/10.1080/19312458.2015.1118446 …
Thanks! Looking more consistent - people’s self-reports of their screen time is untrustworthy and studies that have used self-reports of screen time are probably not actually measuring screen time.pic.twitter.com/t6alfuVaph
it’s an interesting pre-print but I would be loathe to put too much weight on Apple Screen Time data
These results are similar to other studies that have used different ways to objectively assess people’s awareness of their screen time. It’s starting to look fairly consistent that we just stink at self-reporting how much time we spend looking at screens. https://www.tandfonline.com/doi/full/10.1080/19312458.2015.1118446?scroll=top&needAccess=true …pic.twitter.com/rHGKMJSl6b
My point is that Screen Time is a new app, and from personal experience it is unreliable
Sounds like a study waiting to happen.
I suspect that they will soon release technical information, not to mention update the app itself, as more researchers start using it, which will hopefully give more clarity
Nice thread. I’m wondering what you think the findings would be if smartphone use was measured ideally, do you think it would show a more positive or a more negative effect on wellbeing compared to the results so far?
Who knows for sure and that is the probelm. We have focused so many resources on a problem which has no real evidence exists. Would love to see a study comparing objective/subjunctive SM use to outcomes to see if all of our fears were just based on some methodological artifact.
What do you think are the implications for research using the ecological momentary assessment methodology? It’s advantage so far was hugely based on the assumption that these assessments are closer to the „action“. Should that be revised? How could we test that?
EMA measurement is a very much under-researched topic. Items used are often not validated, or validated for retrospective assessment ('in the last 2 weeks') & reworded, which might not translate to EMA. Also unclear what answering an item 100 times does (response shift bias).
Here is one example of this kind of work:https://twitter.com/EeskevanRoekel/status/1068139685854224385 …
Oooooh I want to know the results :)
The study is still in progress! The presenter is not on Twitter I think, but I think others from @ccp_leuven can keep us posted (tagging @InezGermeys @LivveyKirtley)
Awesome, cannot wait :).
Gudrun Eisele was presenting this and is working on ESM methodological studies for her PhD @ccp_leuven. Gudrun is not on twitter yet, but I am doing my best to persuade her! Data collection has not commenced yet, but stay tuned for results!
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
