Akhtar-Danesh, Noori, & Stephen C. Wingreen (2020, November 9). How to analyze change in perception from paired Q-sorts. Communications in Statistics –– Theory and Methods. (ePub in advance of print) (doi: 10.1080/03610926.2020.1845734) (Link: https://doi.org/10.1080/03610926.2020.1845734)
Abstract: Although there have been some previous attempts on analyzing changes in perceptions in Q-methodology, a systematic approach is lacking. In this article we introduce two new methods for analyzing change in perceptions in Q-methodology using paired Q-sorts. We also demonstrate these methods using an actual dataset. Method I: This approach is appropriate for assessing the changes in perceptions between two different conditions of instruction applied to the same subjects. The changes are assessed using a factor analysis on the differences between the Q-sorts from the two conditions of instruction. Method II: This method examines the changes in perception from a baseline Q-analysis. This is usually appropriate when data are collected at two time-points, e.g., before-after situations, where the first assessment is considered as the baseline. In this approach, a by-person factor analysis is conducted on the baseline Q-sorts (condition 1) and factors are identified. Then, the changes in perceptions are assessed for the subjects loaded on any factor from baseline using the Q-sorts from condition 2. In conclusion, these two methods are easy to apply, the results are more objective, and are less prone to investigator bias.
Noori Akhtar-Danesh <daneshn@mcmaster.ca> is in the School of Nursing, McMaster University, Hamilton, Ontario, Canada; and Stephen C Wingreen <stephen.wingreen@canterbury.ac.nz> is in the Department of Accounting and Information Systems, University of Canterbury, Christchurch, New Zealand.
