Conversation

Aella
@Aella_Girl
I asked ~14,000 people if they'd cheated in their relationship, and if their partner had cheated. Here's a graph of (straight) relationships over time, showing the gap between people's actual cheating rates and the average guess of if their partner has cheated. 1/
Image
64
1,075
Neha Hridaya
@neha_hridaya
Replying to
@Aella_Girl
To make any conclusions about any data, you need to first understand your sample selection well. Who are these 14K people and how did you select them? So you can understand the implicit biases in your selection and how they may impact your conclusions.
2
15
Neha Hridaya
@neha_hridaya
Replying to
@neha_hridaya
and
@Aella_Girl
For example, if the men interviewed are only/mostly men who visit sex workers, then duh your sample will show a higher rate of cheating. This is why you need to explain how the sample of 14k was selected and what are their attributes before you can even begin to conclude anything
2
8