Author Topic: Common scale in multidimensional analyses?  (Read 384 times)

Georg II

  • Newbie
  • *
  • Posts: 21
    • View Profile
Common scale in multidimensional analyses?
« on: August 19, 2013, 12:18:24 PM »
Dear colleagues,
Are person and item parameters in a multidimensional analysis in ConQuest on a common scale?
If the mean of persons on dimension 1 is .50 and on dimension 2 is .90 can we conclude that dimension 1 is harder or persons are more proficient on dimension 2?

Cheers
Georg

Eveline Gebhardt

  • Administrator
  • Full Member
  • *****
  • Posts: 103
    • View Profile
    • Email
Re: Common scale in multidimensional analyses?
« Reply #1 on: October 07, 2013, 12:09:40 AM »
The scales of MD models are not comparable that way. By default, the mean of the item difficulties of each scale is fixed to zero. The average performance of students is relative to the average item difficulty of each scale. If you add a few easy items to the second dimension, the average of the item difficulties will be zero again and the average ability will therefore increase. You can compare the scales by saying that the items measuring one dimension are less well targetted (too easy or too hard) than items measuring another dimension. It's not the dimension that is hard or easy, its the items that were selected to measure the dimension.