Jacob Lee
@jacobilee.bsky.social
Research Associate at Fralin Biomedical Research Institute at VTC
Would RFK even get confirmed by the Senate? I sort of doubt it.
November 4, 2024 at 7:04 PM
Would RFK even get confirmed by the Senate? I sort of doubt it.
but if you are writing and advertising your software for others to use, maybe even publishing papers about how great it is, then resist the urge to make your software do more; wait until you know it actually does what you advertise and does it well. Writing good software is HARD. 2/
September 6, 2024 at 12:11 AM
but if you are writing and advertising your software for others to use, maybe even publishing papers about how great it is, then resist the urge to make your software do more; wait until you know it actually does what you advertise and does it well. Writing good software is HARD. 2/
If you don't have the technical chops to pull it off, then hire someone who does. Anything less is just shifting the burden onto your participants, and onto all your professional colleagues who do put in the effort.
9/9
9/9
May 25, 2024 at 2:46 PM
If you don't have the technical chops to pull it off, then hire someone who does. Anything less is just shifting the burden onto your participants, and onto all your professional colleagues who do put in the effort.
9/9
9/9
And, here is the most frustrating part of this. 90% of the time a social psychology experiment uses deception it isn't necessary to the study. It's done because making interactive multi-participant tasks is a technical challenge, or because they haven't thought hard enough about how to do it.
8/n
8/n
May 25, 2024 at 2:43 PM
And, here is the most frustrating part of this. 90% of the time a social psychology experiment uses deception it isn't necessary to the study. It's done because making interactive multi-participant tasks is a technical challenge, or because they haven't thought hard enough about how to do it.
8/n
8/n
So, to sum:
1. You harm participants (however slightly)
2. You harm science
7/n
1. You harm participants (however slightly)
2. You harm science
7/n
May 25, 2024 at 2:41 PM
So, to sum:
1. You harm participants (however slightly)
2. You harm science
7/n
1. You harm participants (however slightly)
2. You harm science
7/n
The one exception is when an experimental economist tells me something. I believe them. Because they too have earned a reputation. Deception is taboo in experimental economics, and I have never personally been debriefed in an economics study where they told me they lied. Just doesn't happen.
6/n
6/n
May 25, 2024 at 2:40 PM
The one exception is when an experimental economist tells me something. I believe them. Because they too have earned a reputation. Deception is taboo in experimental economics, and I have never personally been debriefed in an economics study where they told me they lied. Just doesn't happen.
6/n
6/n
When I am told that I will be interacting with another participant online, I don't believe them. Why should I? The past dozen social psychology experiments where they've told me so debriefed me that they had deceived me. Face it, they've earned a reputation.
5/n
5/n
May 25, 2024 at 2:37 PM
When I am told that I will be interacting with another participant online, I don't believe them. Why should I? The past dozen social psychology experiments where they've told me so debriefed me that they had deceived me. Face it, they've earned a reputation.
5/n
5/n
Social psychology experiments, particularly online ones, are perhaps the worst offenders in this regard. I say so from personal experience participating in studies on Mechanical Turk and Prolific.
4/n
4/n
May 25, 2024 at 2:35 PM
Social psychology experiments, particularly online ones, are perhaps the worst offenders in this regard. I say so from personal experience participating in studies on Mechanical Turk and Prolific.
4/n
4/n
multiple studies, each run independent of the other. This is particularly true on platforms like Mturk & Prolific. If a person participates in one study and they learn in the debriefing that they were deceived, will they believe what the researcher says in the next study? Why should they?
3/n
3/n
May 25, 2024 at 2:31 PM
multiple studies, each run independent of the other. This is particularly true on platforms like Mturk & Prolific. If a person participates in one study and they learn in the debriefing that they were deceived, will they believe what the researcher says in the next study? Why should they?
3/n
3/n
they fail to do their duty of beneficence to participants, and to science itself.
First, deceiving participants actually does harm. No one likes to be lied to, even about the little things. Second, you are not running your study in isolation. Participants often contribute to multiple studies.
2/n
First, deceiving participants actually does harm. No one likes to be lied to, even about the little things. Second, you are not running your study in isolation. Participants often contribute to multiple studies.
2/n
May 25, 2024 at 2:29 PM
they fail to do their duty of beneficence to participants, and to science itself.
First, deceiving participants actually does harm. No one likes to be lied to, even about the little things. Second, you are not running your study in isolation. Participants often contribute to multiple studies.
2/n
First, deceiving participants actually does harm. No one likes to be lied to, even about the little things. Second, you are not running your study in isolation. Participants often contribute to multiple studies.
2/n
pandoc is amazing software.
April 20, 2024 at 7:57 PM
pandoc is amazing software.
Using some kind of static site generator for that? Went that way away from wordpress for my personal site because there really wasn't any need for something like that (e.g. comments)
April 20, 2024 at 3:56 PM
Using some kind of static site generator for that? Went that way away from wordpress for my personal site because there really wasn't any need for something like that (e.g. comments)