I'm not quite sure what you're doing, Jonathan.
Would I be right that you're looking at:
"Needs to be Done (versus) Spends Money"
and calling it two constructs?
This is actually one construct, with an emergent pole "needs to be done"
and an implicit pole "spends money"
Or are you looking at two constructs:
"Needs to be done (versus) *pole unspecified*"
and
"Spends money (versus) *pole unspecified*"
Next, why do you need to ask your respondent to indicate a negative in
order to do an analysis? (Not all constructs have a preferred and
unpreferred pole. Consider the construct
"good social skills - good technical skills" when talking about an
employee. In many contexts neither pole would be preferred over the
other!)
Why are you rating constructs against all the elements? The idea when
doing a grid is to rate elements on constructs and not the other way
round. Or could it be that you're calling the elements constructs and
vice-versa?
>I hope I have not confused
>anyone.
Okay, I own up: I'm confused!
Perhaps if you were to reproduce just a portion of one of your grids- say
just three of the elements and two of the constructs as they're laid out
on the page- your account would be easier to follow. You know:
...........................E1.....E2.......E3
construct 1 left pole.....R11.....R12......R13....... construct 1 right
pole
construct 2 left pole.....R21.....R22......R23....... construct 2 right
pole
Then it would perhaps be easier to answer your question about the factor
analysis.
NB I wish more people would do as you do and involve the respondent in
the naming of factors!
Kind regards,
Devi
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%