Time of day effects on data quality

Hello,

Sorry if this has been addressed elsewhere already.
I am wondering if people have found any time-of-day effects on their data quality?

For example, I have a performance filter for my experiment: the participants need to get 85% to move to the next stage.
I find that participants after 4pm do the best, they have the highest success rate (see the attached breakdown).

Has anyone else observed something similar?

Levan Bokeria

3 Likes

This is a super interesting finding from @Levan_Bokeria

I am wondering if people have found any time-of-day effects on their data quality?

For example, I have a performance filter for my experiment: the participants need to get 85% to move to the next stage.
I find that participants after 4pm do the best, they have the highest success rate (see the attached breakdown).

Has anyone else found anything similar? @trust_level_0

3 Likes

I made exactly the same observation! I had a slightly complex study and the number of people who asked questions which were indicative of not reading the instructions, or failed to complete all parts, was much higher before 4pm (UK time). Since then, I actually only collect data later in the day.

2 Likes

Thanks!
Does anyone at prolific have data on demographic differences by time-of-day? We might decide to only run people after 4pm if they do best in the study, but we don’t want to bias the population too much.

In my data, I tried to look at demographic differences (like age) but the data is too noisy to say anything. I haven’t ran enough participants per time-bin.

1 Like

This is super interesting. Our data team is looking into this, so any other thoughts/theories/data points that you have will be very useful for our investigation.

Btw, our latest #TipTuesdays post covers how you can monitor/improve participant attentiveness

2 Likes