IMO it is hard know what inference to draw from these comparisons.
Firstly, making multiple comparisons obviously raises the risk of a “false-positive” … a result that is merely due to chance/sampling.
Secondly, with ‘multiple hurdles’ it’s hard to know how to compare like for like....
The share of not highly engaged non-males which had ‘personal connection’ as an important factor for involvement was slightly higher than the male counterpart
--> But note that the involvement factors may be driving engagement itself, and doing so differently for males and females
IMO it is hard know what inference to draw from these comparisons.
Firstly, making multiple comparisons obviously raises the risk of a “false-positive” … a result that is merely due to chance/sampling.
Secondly, with ‘multiple hurdles’ it’s hard to know how to compare like for like....
--> But note that the involvement factors may be driving engagement itself, and doing so differently for males and females